<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Bjorn Vuylsteker on Medium]]></title>
        <description><![CDATA[Stories by Bjorn Vuylsteker on Medium]]></description>
        <link>https://medium.com/@Tyzer34?source=rss-75cde01c7b82------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*UcJKxnjb0w6SWILKy_19qQ.png</url>
            <title>Stories by Bjorn Vuylsteker on Medium</title>
            <link>https://medium.com/@Tyzer34?source=rss-75cde01c7b82------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Mon, 16 Nov 2026 14:28:03 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@Tyzer34/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Exploring the World of Generative AI and Large Language Models: Some Personal Notes]]></title>
            <link>https://medium.com/@Tyzer34/exploring-the-world-of-generative-ai-and-large-language-models-some-personal-notes-4d79cec87c58?source=rss-75cde01c7b82------2</link>
            <guid isPermaLink="false">https://medium.com/p/4d79cec87c58</guid>
            <category><![CDATA[product-management]]></category>
            <category><![CDATA[chatgpt]]></category>
            <category><![CDATA[large-language-models]]></category>
            <category><![CDATA[generative-ai]]></category>
            <category><![CDATA[entrepreneurship]]></category>
            <dc:creator><![CDATA[Bjorn Vuylsteker]]></dc:creator>
            <pubDate>Tue, 28 Feb 2023 07:34:55 GMT</pubDate>
            <atom:updated>2023-02-28T07:34:55.797Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*8u-a84qP69_mrmGJcwfCsg.png" /><figcaption>A robotic figure wearing a futuristic augmented reality headset stands in a bustling cityscape with neon lights and towering skyscrapers.</figcaption></figure><p>I have always been fascinated by the potential of Generative AI and Large Language Models. Working in Product at WeGroup, I have seen first-hand how these technologies can revolutionize the way engineers, product managers and content writers work. Through my experience, I have come to understand the importance of security, cost, scalability, and the time it takes to develop and launch applications when utilizing such technology. In this post, I will be sharing some personal notes and thoughts on the world of Generative AI and Large Language Models, and how businesses can leverage these technologies to create powerful applications and content.</p><h3>Consider Security and Cost When Utilizing Generative AI and Large Language Models</h3><p>Generative AI, GPT3, ChatGPT and Large Language Models are powerful tools that can help businesses save time, become more efficient and structure their content. However, when using such technology, it is important to consider the security of sensitive customer data and the cost of using Open AI’s API to generate words. Security is especially important when dealing with customer data as ChatGPT could be using your input and prompts to train new models on. Additionally, businesses must consider cost as using Open AI’s API to generate words can be expensive.</p><p>Furthermore, it is important to be aware of the technical limitations of the technology. For example, there is usually a maximum number of characters allowed to be inputted. Product managers must take this into account when designing the application and consider the boundaries of what’s possible with such technology. It is also important to think about scalability and the time it takes to develop and launch the application to ensure the product is successful. Considering alternative open source solutions such as <a href="https://devforth.io/blog/gpt-j-is-a-self-hosted-open-source-analog-of-gpt-3-how-to-run-in-docker/">GPT-J</a> could really benefit your business use-case.</p><blockquote>“Security and cost are two of the most important considerations when building products with Generative AI and Large Language Models”</blockquote><h4>Tips and Tricks:</h4><p><em>1. When utilizing Generative AI and Large Language Models, ensure that customer data is secure and the cost of using Open AI’s API to generate words is taken into account.<br>2. Understand the technical limitations of the technology, such as the maximum number of characters allowed to be inputted, to avoid any unexpected issues.</em></p><h3>Exploring the Power of Large Language Models and Their Ability to Process Data</h3><p>The power of large language models lies in their ability to process unstructured and structured data. This allows them to generate readable content that is applicable to a wide audience. Additionally, they can be used to generate content for social media, invitations and landing pages. This is especially useful for businesses looking to create content quickly and efficiently.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*TsqfV8JM9Rn2hZFECte3Ag.png" /><figcaption>A robotic figure stands in the center of a bustling city, surrounded by giant holographic screens displaying content generated by large language models.</figcaption></figure><p>Large language models can also be used to create personalized content. By creating a familiar format where users can recognize themselves and configuring the tool, businesses can differentiate themselves from the competition. However, it is important to be aware of the limitations of ChatGPT and to exploit it’s strenghts in generation and summarization. Businesses must start by understanding the problem they are trying to solve for, before they go looking for a new shiny solution utilizing ChatGPT.</p><h4>Tips and Tricks:</h4><p><em>1. Utilize Large Language Models to generate content for social media, invitations, and landing pages quickly and efficiently.<br>2. Create a familiar format where users can recognize themselves and configure the tool to differentiate from the competition.<br>3. Understand the problem and look for the best solution before falling for the </em><a href="https://everhour.com/blog/shiny-object-syndrome/"><em>Shiny Object Syndrome</em></a><em>.</em></p><h3>Creating a Strong Identity and Using Variables Correctly for Content Creation</h3><p>When creating content using Generative AI, it is important to consider the identity of the model. Giving the model an identity allows it to make more accurate decisions and create content that is more consistent with the identity. Additionally, when writing <a href="https://neuroflash.com/chatgpt-how-to-write-the-perfect-prompts/">prompts</a>, it is important to consider the variables and variable names. This ensures that the system understands the relationships between the variables and can accurately generate content. By creating a strong identity and using variables correctly, businesses can generate high-quality content quickly and efficiently.</p><h4>Tips and Tricks:</h4><p><em>1. Consider the identity of the model when creating content using Generative AI to make more accurate decisions and create content that is more consistent.<br>2. Use variables correctly when writing prompts to ensure the system understands the relationships between the variables.<br>3. Create a </em><a href="https://prmpts.ai/"><em>prompt library</em></a><em> to quickly reuse prompts and speed up the content creation process.</em></p><h3>Leveraging AI to Innovate, Collaborate and Create</h3><p>It is important not to think only in terms of GPT or Large Language Models, but also to think of the collaboration with other artificial intelligence tools. For example, leveraging AI to create end-to-end solutions will prove to be more efficient, scalable and personalized. This can help businesses save time and make the effective results more understandable for and loveable by the end user. A great example of this can be found <a href="https://www.linkedin.com/feed/update/urn:li:activity:7034434559685406720?updateEntityUrn=urn%3Ali%3Afs_feedUpdate%3A%28V2%2Curn%3Ali%3Aactivity%3A7034434559685406720%29">here</a>.</p><blockquote>“Leveraging AI to collaborate and create can help businesses save time and make the information more understandable for the end user”</blockquote><h4>Tips and Tricks:</h4><p><em>1. Think of the collaboration with other AI tools to create a content planning that is more efficient, scalable, and personalized.<br>2. Look at the impact on people’s lives to understand the value of the product.<br>3. Consider the cost of using AI tools and the time needed to develop and launch the application.</em></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*RrIyjApbIQkRo4y73W6HWQ.png" /><figcaption>Post apocalyptic city scape with a hive mind hovering over the city, tubing humans and taking their lifeforce energy.</figcaption></figure><p>Ultimately, Generative AI, GPT3, ChatGPT and Large Language Models do help businesses create powerful applications and content. When using such technology, it is important to consider security, cost, scalability, and the time needed to develop and launch the application. Leveraging AI to create, collaborate and inspire can also be beneficial. Additionally, businesses should use a strong identity and define variables correctly when creating prompts to ensure high quality and efficient content is produced. By understanding the various features and limitations of Generative AI, GPT3, ChatGPT and Large Language Models, businesses can grow and easily outpace their dormant competition.</p><p>My journey of exploring the world of Generative AI and Large Language Models has been an exciting one so far, and I am sure that your business can benefit greatly from these technologies. If you have any questions about GPT and Large Language Models, I am always open to discussing use cases or grabbing a coffee. I hope that this blog post has been helpful to you, and I look forward to hearing your thoughts and experiences on this new amazing era we live in.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=4d79cec87c58" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[“Louise, I have a problem.”]]></title>
            <link>https://medium.com/wegroup/louise-i-have-a-problem-af5f42f318f0?source=rss-75cde01c7b82------2</link>
            <guid isPermaLink="false">https://medium.com/p/af5f42f318f0</guid>
            <category><![CDATA[insurance]]></category>
            <category><![CDATA[chatbots]]></category>
            <category><![CDATA[insurtech]]></category>
            <category><![CDATA[claims]]></category>
            <category><![CDATA[machine-learning]]></category>
            <dc:creator><![CDATA[Bjorn Vuylsteker]]></dc:creator>
            <pubDate>Sun, 27 Jan 2019 13:59:06 GMT</pubDate>
            <atom:updated>2019-01-28T17:35:49.187Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1000/0*bTBJHJKhXZeqW5wt" /></figure><p>If you have heard of WeGroup, chances are that you have heard of Louise, our virtual insurance assistant! She represents the front of WeGroup technology, which allows her to assist insurance providers with analysing consumer’s insurance need, setting up policy propositions, automating back-office work, but also providing help to work towards claims automation.</p><blockquote>Automatic claims handling enables claims managers to spend more time on the difficult and intricate cases.</blockquote><p>Recently, I got the chance to talk about the latter— claims automation — at the 8th Annual Insurance Claims Management Conference in Lisbon.</p><p>Personally, I am not a big believer of the concept of <em>robots are taking over our jobs</em>. I am into technology enabling us to be better at what we do, and that is exactly what WeGroup is destined to do. It enables a claims manager, for example, to be able to spend more time on the difficult and intricate cases, and as such attaining higher customer satisfaction, while the affairs that have become routine are automated.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*CbaItY1UimLWb5TyXKQbXQ.png" /></figure><p>We all know that computers are great at working with standardized and structured data like spreadsheets, or database tables. They are able to process data much faster than humans. However, we do not communicate in <em>structured data</em>. We communicate using words, which is a form of unstructured data. The problem here is that a simple computer program might be able to capture the words, but it won’t know what to do with them. It will understand the syntactics, but not the semantics. And this makes language understanding a rather difficult problem. Similar words might have different meanings and different words might have the same meaning. So, in order to understand the meaning behind a word, you must always explore it in its context.<br><br>This is exactly what <strong>Natural Language Processing</strong> (or NLP) does. It aims at exploring the given text or sentences and relating words to each other in their context. This way, the computer will have a better understanding of the structure of the text, its entities and its intents. Eventually, NLP can be seen as a tool that helps a computer understand our unstructured data, and transform it into structured data.</p><p>In conversation with a chatbot or virtual assistant, the computer should be able to understand the intent of the user and reply accordingly. However, this technology is not exclusive to chit-chat conversation.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*PYJlj4lYh7NWY3BA4kOnwQ.png" /></figure><p>Within WeGroup, we use NLP technology to enhance our property claims management. First, we ask ourselves the question, what information do we actually need to submit a property claim? Well, first of all, we need to know what type of risk occurred. Was it caused by leakage? A small fire or burglary perhaps? Secondly, we should know what objects were damaged or taken. Third of all, we should know when this happened with a specific time and date. Next, to that, it is also very interesting to be able to understand what caused the incident and in what area of the property it occurred.<br> <br>Now the question is, as speech is brought in an unstructured form, how would a computer be able to extract these specific key information from their context? Using NLP technology, we were able to build a system that takes text as an input and gives back the found key information as a result. Coupled with our Louise, we can then dive deeper into the unanswered questions, to extrapolate the key information needed from the customer. Let’s give an example of how we teach a computer to extract this information.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*lD2iFkHNJJBGhw7CSbZnlA.png" /></figure><p>One of the key features needed is the date of occurrence. Extracting time from a textual context is a relatively simple problem, as there are only so many ways one can tell the time. For example, you could type: “<em>It happened Wednesday, at five past half two in the afternoon.</em>” or “<em>It happened the 23rd of January at around 1:35 pm.</em>” or as it is text based, in a more digitally structured way.</p><p>However, other information such as extracting the type of risk occurred is a quite difficult problem. To solve this, we use a machine learning algorithm that predicts the type of risk based on certain word pairs extracted from the text. The algorithm relates the words “kitchenrobot” and “exploded” in a sentence, which implies a fire-related risk. If it were to extrapolate the word pair “bathroom” and “wet”, this would imply a water-related risk.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*AyhVKrYuXwNVRx0_ncM6RQ.png" /></figure><p>We can even take this one step further. Because we now have our claim in a structured format, we can check two big flags for approval. That is, first, if the claim is <strong>valid</strong> under its given policy, and secondly, if the claim is not deemed <strong>fraudulent</strong>.</p><p>In order to check if the claim is valid under its given policy, we use NLP technology to analyse the policy terms of a given policy to extract what is actually covered and what is not. In doing so, we are able to check the claim against the policy terms and quickly conclude whether or not the damage that occurred is deemed covered. For example, if the claim concerns water ingress and an open window and we check it against a given property policy, we can assess that the claim is deemed invalid as this occurrence is not covered.</p><p>Secondly, once we know that the claim is deemed as covered by its policy, we must check for fraudulent activity. As the claim is already given in a structured form, it is much more straightforward to check for inconsistencies. These inconsistencies are mostly based on factual information in relation to what could be deemed fraudulent. However, as enough use-cases are gathered, machine learning algorithms are able to find unseen anomalies, from which the system learns again.<br> <br>At WeGroup, we do not only aim at building technology to enhance the current way of working, but we also work towards future proof solutions. I have shown you that current technologies enable us to bring another way of handling insurance claims. However, with this technology, we can also tackle the products of tomorrow. We truly believe that, in the future, many of the digital applications that we know today will be enabled with voice technology via home assistants such as a Google Home or Amazon Alexa. These devices allow us to capture speech in a new, fun and exciting way. However, that speech is then converted to text, which implies that we are able to attach our Louise enabled with NLP technology to these devices.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/800/1*HAk24zJQ7cnGDkaY9bDq9g.gif" /></figure><p>Let’s say that someone wants to file a claim with a Louise device. The person will say: “<em>Hey Louise, I want to file a claim.</em>”. Starting from that point on, a conversation will be started with Louise, who will answer from the Alexa: “<em>Hey Bjorn! Hopefully, you are alright? What happened?</em>”. The user will then explain what happened and from that, Louise will once again extract the key information. If some information is unknown to her, she will ask directly for that information and the user can reply to the given question accordingly. However, this goes beyond claims automation. Imagine having bought an electric bike and wondering if your insurance actually covers liability for it. One could ask Louise: “<em>Hey Louise, am I covered for liability in regards to my electrical bike?</em>”. Hopefully, she will respond that you do, but in the case you don’t, she could recommend you the optimal insurance policy that does cover this specific part and insure you on the spot.<br> <br>So basically, insurance should not be seen as a bothersome activity, but should be integrated within our daily routines. This is why we truly believe that using the technology of today, we can enhance processes from the past, in order to build a better future and <em>make insurance great again</em>!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=af5f42f318f0" width="1" height="1" alt=""><hr><p><a href="https://medium.com/wegroup/louise-i-have-a-problem-af5f42f318f0">“Louise, I have a problem.”</a> was originally published in <a href="https://medium.com/wegroup">WeGroup</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The Fairy Tale Concept of Happiness — The Perception of a 22-year-old]]></title>
            <link>https://medium.com/@Tyzer34/the-fairy-tale-concept-of-happiness-the-perception-of-a-22-year-old-d6b153c0414?source=rss-75cde01c7b82------2</link>
            <guid isPermaLink="false">https://medium.com/p/d6b153c0414</guid>
            <category><![CDATA[friends]]></category>
            <category><![CDATA[opinion]]></category>
            <category><![CDATA[happiness]]></category>
            <category><![CDATA[life-lessons]]></category>
            <category><![CDATA[family]]></category>
            <dc:creator><![CDATA[Bjorn Vuylsteker]]></dc:creator>
            <pubDate>Sat, 29 Jul 2017 10:51:50 GMT</pubDate>
            <atom:updated>2022-12-19T07:50:22.772Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*JIkUKIg6TK-GTxNcH5Ad7A.jpeg" /></figure><p>In all certainty, I am not one to be quick to write a piece reflecting my opinion on certain matters. I’d rather stay on the background and silently form my own opinion based on facts and research. However, with this particular topic, I don’t really feel like background research is at hand here. Everyone has their own opinion on what happiness is and what it means to them and others around them. To be honest, as a 22-year-old, I may not have gained enough life experience to touch on this topic. However, I think that maybe seeing things from my perspective could prove to be rather refreshing and you might reflect on your own opinion as to what happiness means to you. Actually, that is what I am trying to do with this article. I am not one to make a statement on how to be happy, but I am just trying to reflect on the things in life that bring me happiness and want to invoke a discussion on why you might think differently.</p><p>For me, happiness consists of many small pieces, and during this article, I will try to explain why each piece contributes towards my happiness. An important one is <strong>freedom</strong>. Being able to do the things you love and want to do. The freedom to express yourself, whether it is via music, art or dancing. The freedom to make your own choices in life and to walk your own path. I am privileged to have been given the chances in life to pursue playing music, to attend university and to choose my studies without any external pressure. This freedom has led me to explore many things in my own peculiar way and has enabled me to formulate my own opinions about many different topics.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*tbbpIZGFB2jrA2G1HBwEFA.jpeg" /><figcaption>Our delegation of the first ever Belgian Huawei Telecom Seeds for the Future in China [2014]</figcaption></figure><p>One thing that contributed to my happiness early on, is that I quickly learned the act of <strong>initiative</strong>. At the time, I had no idea how much impact it would have on my life. Initiative led to social interactions, which in its turn allowed me to become a people person. This was crucial as I learned how to converse with all different types of people. It also allowed me to grab many of the opportunities that presented themselves during my education. Because of this, I was able to traverse to China for three weeks with Huawei.</p><p>I also believe that, if life doesn’t give you many opportunities, you have to <strong>create them yourself</strong>. For example, it is not because you didn’t take up coding classes in university, that you’ll never be able to work as a developer. Take matters into your own hands, take some online courses, earn certificates, make some hobby projects to help you code and show those coding companies that you are really motivated to be a developer. Make your dreams come true and never let anyone tell you otherwise.</p><p><em>Hakuna Matata</em> — <strong>Don’t worry, be happy.</strong> It is a slogan that I live by as an everlasting optimist. I learned that worrying about the small things in life only leads to stress, but doesn’t actually change anything about the outcome. Why would you stress about it then? Take the following example. You are a student and have just finished your exams. Many people would start stressing out about the results, even if it is still weeks to come before the results are known. But why bother? The exams are submitted anyways and there is no possible way that worrying about the result will have any sort of impact on the end result! By that reasoning, I’d rather enjoy my vacation period without worrying or stressing about school.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*mJMt-eXF4Fk8cW7Lxs2byA.jpeg" /><figcaption>With my colleagues at Kinepolis Kortrijk, we took a trip to The Ardennes to bond. [2016]</figcaption></figure><p>Happiness is also something that I believe to be very <strong>contagious</strong> and I try to spread it to the fullest. I feel happy when my environment is feeling happy and if they are not, they know that they have someone to talk to because I do feel better knowing my environment is happy.</p><p>Sharing happiness also means that you should <strong>be happy for others</strong> and their accomplishments. Don’t be jealous, but accept the fact that they achieved something and be happy for them. In the case you were contesting for the same position, for example, give the other the respect he or she deserves and aim even higher, because what is life without a challenge?</p><p>With that said, I also love turning life itself into one big challenge. Those dishes you need to do? Measure how long they take and try to beat your best time. If you keep on making challenges, you might end up surprising yourself with what you are capable of.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ZM_iI42Tf8RM0nFARD8JyA.jpeg" /><figcaption>Pohoda Festival in Slovakia, one of the best festivals I’ve ever done. [2017]</figcaption></figure><p><strong>Friends and family</strong> are also certain to play a big role in my happiness. Think back to some of the moments during the last year when you were truly happy. For me, during these moments, there were always people besides me that shared these moments with me. Not only did they elevate the moments themselves, they also helped create the memory and make it memorable for many years to come.</p><p>For most people, success leads to happiness, and why wouldn’t it? Who doesn’t feel happy when they succeed at something they have worked a great deal for? For me though, <strong>failure is also key to happiness</strong>. I am a very driven person and am always looking for ways to improve myself, and that is exactly what failure brings about. It shows me a way to improve myself, so I can learn from my mistakes and succeed the next time.</p><p>Something I don’t believe in though is the concept of finding happiness. People should be able to create their own happiness around things they feel happy about. Even through the toughest times, there will always be moments that bring about happiness, you only have to embrace those moments and focus on the positive instead of the negative. Trust your friends and rely on them when you need them because they will be there for you as you are for them.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*xSk4d76RJrI2_ITaAYipZA.jpeg" /><figcaption>A view over Budapest from The Citadel. I could have spent hours here. [2017]</figcaption></figure><p>Start living <strong>the life of an optimist</strong>. It is so much more fun! Even when you are in a pinch, deadlines have to be met, reports have to be filed, etc. You have to stay optimistic and think about the fun stuff you will be able to do when you are done. One thing that helps me do this, is careful planning and organisation. I love to plan ahead, whether it is schoolwork, city trips or even regular weeks. Because of careful planning, I know that I do not have to worry about deadlines and reports being filed. They will always be done in time, as long as I stick to the plan. This is also one of the main reasons that I can remain stress-free during very busy periods. It also enables me to enjoy the real moments to their fullest. Take a city trip for example. Because everything is planned out in advance, I do not have to worry about catching the plane, making reservations in time, etc. because I know that my planning is accurate. This allows me to enjoy the city trip to its fullest.</p><p>My happiness is also not based on the approval of others. In this digital age, there are many people that go out, take pictures for social media and wait. Wait for the likes on Facebook, the hearts on Instagram, basically wait on the approval of others. If the picture gets a lot of likes? Great! If it does not, it is a disaster. People start questioning the things they are doing, the way they are dressed, the meaning of life itself! Maybe I am over dramatizing here, but I do want to make a point that you should live the life you love and stop caring about other peoples opinions. You should <strong>do things for your own reasons, and not because others will approve of it</strong>. Stop living on the internet and start to really live. Enjoy moments when they present themselves. For example, I love going to concerts. However, I am always bothered by people that film the whole concert with their smartphone. Why would you come to a concert then? You could have watched it at home on a live stream! I do have nothing against taking a few pictures or filming a small part, but a whole concert, <em>incomprehensible</em>!</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*W-flYIOvNNfZcOUmcYbLbQ.jpeg" /><figcaption>Sunset dinner by the water in Venice. Those are the moments that make life worthwhile. [2016]</figcaption></figure><p>For me, the <strong>small things in life</strong> also contribute greatly towards my happiness. Mowing the grass barefooted, lying in the sun, dreaming away with headphones on, going out for dinner, enjoying wonderful views on vacation. There are so many little moments in life where I can honestly say to myself: “<em>Now I am happy</em>”. Cherish these moments and enjoy them while they last.</p><p>Last but not least, happiness for me is the fact that I know that I still have a full life ahead of me. That I am still able to explore many new places, meet many wonderful people, learn many new things, live up to my own expectations and maybe enjoy the wonders of starting a family. Maybe I am just a naive 22-year-old who still thinks of happiness as a fairy tale concept. Maybe I am not, and my way of looking at life is refreshing for some. Maybe I’ll change my opinion in a few years, who knows? All I know is that happiness for me is here and that I am having the time of my life.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=d6b153c0414" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Speech Recognition — A comparison of popular services in EN and NL]]></title>
            <link>https://medium.com/craftworkz/speech-recognition-a-comparison-of-popular-services-in-en-and-nl-67a3e1b0cee6?source=rss-75cde01c7b82------2</link>
            <guid isPermaLink="false">https://medium.com/p/67a3e1b0cee6</guid>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[chatbots]]></category>
            <category><![CDATA[speech-recognition]]></category>
            <category><![CDATA[python]]></category>
            <category><![CDATA[machine-learning]]></category>
            <dc:creator><![CDATA[Bjorn Vuylsteker]]></dc:creator>
            <pubDate>Mon, 13 Feb 2017 07:51:19 GMT</pubDate>
            <atom:updated>2017-02-13T07:51:19.811Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*jwGgu6FYFLqLofixEN_Xow.png" /></figure><p>With the rise of robotics and voice-enabled services, Automatic Speech Recognition (ASR) — or Speech-To-Text (STT) — is more important than ever. It can be used to enable conversation with a chatbot or robot, or to automatically transcript conversations or movie scenes. Overall, it is not a fairly new field, but a challenging one to say the least.</p><p>The concept of speech recognition goes way back to the 60’s, but as a Belgian, it all began in the late 80’s, when the infamous company named <em>Lernout &amp; Hauspie </em>settled their roots in the Belgian town of Ypres. Over the course of the 15 years, they developed cutting-edge technologies and pushed the field of speech recognition and synthesis to a new level. In 1999, they had the pleasure of opening a technological campus in Ypres, called the <em>Flanders Language Valley</em>, modeled after Silicon Valley. Unfortunately, due to legal complications, Lernout &amp; Hauspie went bankrupt in 2001. It was then that the American company Scansoft — now known as Nuance — took over their technological department.</p><p>Today, Nuance is known as the major runner-up for industrial and medical speech recognition services. However, more and more companies like Google, Apple, Amazon and IBM are catching up with services directed mostly towards the consumers and B2B market. As I was curious to see how services from different companies handled speech recognition, I tested out a few of them and made a comparison for both the English and Dutch language.</p><h3>The Services</h3><p>Although there are a fair amount of platforms out there providing speech recognition services, we will compare three of the (currently) most popular — and easily accessible — ASR services.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*8J8yEexLEtXo7gI2GGEPoA.png" /></figure><p>Known to everyone, the go-to company for translation also provides speech recognition and synthesis services. Although still in Beta version, the Google Speech API provides a huge variety of 40 different languages to recognise speech from. They also present good and structured documentation and offer very easily deployable SDK’s for their services.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*zTeD2-Xb1rSFQbsAcFloqw.png" /></figure><p>As previously mentioned, the American Nuance is a top-notch company in the field of speech recognition and synthesis. Coming forth with an already huge amount of industry specific solutions, they recently launched <em>Dragon Drive</em>, which is already implemented in some BMW vehicles. The service that we are using, however, is the <em>Nuance Mix</em> service. Launched at the end of 2015, the platform comes forth as the IoT and developer solution for their speech services, currently providing 40+ different languages for speech-to-text analysis.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*TCPde83kvLmMwF4_WON9-g.png" /></figure><p>Initially known as IBM’s Jeopardy winning AI service, Watson came forth for commercial applications at the beginning of 2013. With the growth of cloud and IoT, <em>IBM Bluemix</em> was launched as the go-to platform for all related services. Although speech analysis capabilities were only added at the beginning of 2015, early research on ASR at IBM dates all the way back to the 60’s. Currently, Watson’s speech-to-text services provide transcriptions in 9 different languages, although not including Dutch.</p><h3>The Data</h3><p>As the purpose is to compare the previously mentioned ASR platforms, a generic method is set up to test the different aspects of the audio analysis services. First, a 3–4 minute video from a movie or tv show is selected in the preferred language to be analysed. Next, the audio is split into multiple pieces, so that each piece can be analysed separately. Lastly, the textual output and latency of the responses are measured for comparison. In order to implement the audio analysis services, Python was used for traffic to and from HTTP endpoints and for latency measurement.</p><p>Important to note is the difference between the analysis of single audio files — <em>a.k.a. what we will be doing</em> — and audio streams. With audio streams, speech-to-text analysis tends to be on the fly, as it has to perform in real-time and no future data is accessible yet. However, when a service has to analyse a single audio file, it has to make the trade-off between accuracy and latency. This trade-off will become very clear in the results further on.</p><h4>Request</h4><p>In a comparison, it is important that the same input data is used for each service, so an honest result can be achieved. This is done by making sure that each service supports that given input data. Practically, this means taking the audio from the chosen scene, resampling it to 16 kHz and refactoring it from stereo to mono sound.</p><p>Afterwards, the audio fragment is split up into several pieces, as not all services are able to analyse 3–4 minute long audio data. At the start of the audio file, splits are made after a single sentence, so that the ASR services have a higher probability of achieving good results. However, longer cuts are used towards the end of the audio file to make for a more challenging analysis.</p><h4>Response</h4><p>When a service is ready analysing the send audio file, it will return a certain response. This can either be the resulting speech-to-text output or an erroneous response. In the case of the erroneous response, the service could either be unable to identify any speech from the given audio file or a timeout could have occurred.</p><p>Independent from the content of the returned response, the latency of the transaction is measured as the difference in time between the arrival of the response and the creation of the request. Needless to say, latency plays a big part in the consumers domain and even more in real-time speech applications.</p><h3>Comparison with en_US (English)</h3><p>Firstly, we will compare a movie scene in English, taken from the award-winning movie <em>The Wolf of Wall Street</em>. It is a scene where Jordan (Leonardo DiCaprio) gives an outstanding motivational speech to his employees. The tricky part of this movie fragment is that, although most of the speech is very comprehensible, there are parts in between where shouting occurs and words are rather screamed than said.</p><h4>The audio file</h4><p>The chosen movie scene accounts for a 3 and a half minute long audio file, that is split up into 21 several audio fragments as seen below. At first, splits occur at small pauses, making the first audio fragments rather small. Afterwards, longer cuts are used to make sure the mentioned ASR services are tested on both short and long transcriptions.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*r_3_dXt3UAlHyF4DFvT4Ww.png" /><figcaption>The waveform for the audio fragment taken from The Wolf of Wall Street. Audio splits occur at the green lines.</figcaption></figure><h4>Side-by-side comparison</h4><p>If I were to be lazy, the results could be shown in a table, with textual output and some statistical comparison measurements. But why not make it a little bit more fun? In the following video, a side-by-side comparison is shown between the three mentioned ASR services, showing their textual output at that moment of the scene and their measured latency to get that output. Spoiler alert, some of the transcripts might prove to be funny..</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FD3shWnA6cT8%3Ffeature%3Doembed&amp;url=http%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DD3shWnA6cT8&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2FD3shWnA6cT8%2Fhqdefault.jpg&amp;key=d04bfffea46d4aeda930ec88cc64b87c&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/b50b8aeb06c89d7aae1280be51cdfff7/href">https://medium.com/media/b50b8aeb06c89d7aae1280be51cdfff7/href</a></iframe><h4>Results</h4><p>Even at first glance, the fact that there is a difference between these three services is very clear. Looking at the shorter cuts, the Google Speech API has its moments where a good result is achieved, but mostly at the cost of high latency. Watson STT brings an opposite view towards this, having very low latency at the cost of a poor prediction accuracy. Out of these three services, Nuance Mix comes forth as the winner, having both quite low latency and a very high prediction accuracy.</p><p>This becomes even more clear when looking at the longer cuts. Google’s speech services — although still a Beta — take a huge amount of time to come up with literally nothing in return. Watson services still work towards low latency, but produce only a small fraction of the desirable results. Nuance however, is overall able to get a high prediction accuracy at an acceptable latency level, given the length of the input data.</p><p>We can thus clearly conclude that, for this movie fragment and between these services, Nuance Mix is the best option by far.</p><h3>Comparison with <em>nl_NL</em> (Dutch)</h3><p>Although not quite the most common language, I also wanted to analyse how the services would perform on my native language. Unfortunately, this meant that Watson already had to kick the bucket, as of now it does not provide support for Dutch STT analysis. We, however, still have two services left to compare.</p><p>To make it a little bit more challenging, a scene was taken from the popular Belgian television show called <em>De Ideale Wereld</em>, which does not always include the most proper — but still understandable — dutch vocabulary.</p><h4>The audio file</h4><p>The scene to be analysed takes up just a little bit more than 3 and a half minutes and is cut into 29 parts in total. As with the previous comparison, cuts tend to be short at the beginning and longer at the end of the audio file.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*w4LVvIpuqABcxAVWrxGX8g.png" /><figcaption>The waveform for the audio fragment taken from De Ideale Wereld. Audio splits occur at the green lines.</figcaption></figure><h4>Side-by-side comparison</h4><p>Once again, the comparison results of both services were reworked in a video, clearly denoting the difference between them at several points in time.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FUbTZL-i294c%3Ffeature%3Doembed&amp;url=http%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DUbTZL-i294c&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2FUbTZL-i294c%2Fhqdefault.jpg&amp;key=d04bfffea46d4aeda930ec88cc64b87c&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/09ca661b36dbe2521b2e7a53e42495f0/href">https://medium.com/media/09ca661b36dbe2521b2e7a53e42495f0/href</a></iframe><h4>Results</h4><p>The Google Speech API tends to score quite good at the short audio fragments, as does Nuance Mix. The big difference between both services, however, is that on average, the Google Speech API has about double the latency of the analysis using Nuance Mix. This huge difference in latency stays at the longer fragments, with the Google Speech API not even producing any results at certain fragments. Although Nuance Mix leaves out several parts of the text as well, I think its results are quite remarkable, given the fact that the longer audio fragments have a sort of chaotic conversation going on, using words that are not always found in a Dutch dictionary.</p><h3>Conclusion</h3><p>Although speech-to-text analysis has not been perfected yet, the results of both comparisons were above my expectations, especially because I did not use the most common and easy examples.</p><p>Looking at the Watson STT service, it seems clear that IBM strives towards a faster service, trading in accuracy for a shorter latency. This affects the results greatly in prospect to the other services, although it has the certainty of delivering a response in a short time-frame. It is a pity that Dutch is not available yet, but it should be added in the near future.</p><p>Although still in Beta, the Google Speech API produces quite remarkable transcriptions for short audio fragments, at the cost of a high latency. It is however in both the English and Dutch comparison not really able to analyse longer audio cuts, even though the analysis time is quite long compared to the other services. Hopefully, this will be resolved at release.</p><p>Building forth on Belgian technology, Nuance comes up as the true victor of this comparison. It is able to deliver remarkably accurate results, in a very acceptable time-frame. In both comparisons, it is clear that over the years, Nuance has gained the most experience in the field of speech analysis.</p><p>Overall, it was fun setting up a comparison between these three services and I was really surprised with the results Nuance’ services produced. It seems that for now, the choice of service is very application dependent, mostly leaning towards a trade-off between latency and accuracy.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=67a3e1b0cee6" width="1" height="1" alt=""><hr><p><a href="https://medium.com/craftworkz/speech-recognition-a-comparison-of-popular-services-in-en-and-nl-67a3e1b0cee6">Speech Recognition — A comparison of popular services in EN and NL</a> was originally published in <a href="https://medium.com/craftworkz">Craftworkz</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Plex ♥ Alexa — The perfect wedding]]></title>
            <link>https://medium.com/hackernoon/plex-alexa-the-perfect-wedding-38b14b41faf0?source=rss-75cde01c7b82------2</link>
            <guid isPermaLink="false">https://medium.com/p/38b14b41faf0</guid>
            <category><![CDATA[alexa]]></category>
            <category><![CDATA[flask]]></category>
            <category><![CDATA[heroku]]></category>
            <category><![CDATA[plex]]></category>
            <category><![CDATA[flask-python-appengine]]></category>
            <dc:creator><![CDATA[Bjorn Vuylsteker]]></dc:creator>
            <pubDate>Mon, 02 Jan 2017 15:56:43 GMT</pubDate>
            <atom:updated>2019-05-21T09:02:22.563Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*2_JcGOV4g03DL5HF1EhhMA.png" /></figure><p>Only a small two weeks ago, I acquired an Amazon Echo Dot, together with a Bose SoundTouch 10. An amazing combination for music playback! However, few possibilities for music services are currently available. Standard support for Amazon Music Unlimited and Spotify is provided. There is also a skill available for music playback from SoundCloud, called <a href="https://www.alexaskillstore.com/Magic-Jukebox/40064"><em>Magic Jukebox</em></a> and a developer called Steve Gattuso is actively updating his <a href="https://github.com/stevenleeg/geemusic"><em>geemusic</em></a> project, enabling playback from <a href="https://hackernoon.com/tagged/google">Google</a> Music.</p><p>As a developer, I wanted to broaden the spectrum a bit by giving Alexa users access to their own Plex Music Player. As I previously wrote an in <a href="https://blog.craftworkz.co/flask-ask-a-tutorial-on-a-simple-and-easy-way-to-build-complex-alexa-skills-426a6b3ff8bc#.7kl2pqiol">depth tutorial</a> on how to create your own Alexa Skills, I will not be going into much of the details here. This time, <a href="https://hackernoon.com/tagged/heroku">Heroku</a> is used for the deployment of the application, but almost any platform can be used. Do keep in mind that deployment on PythonAnywhere will not work for this application, as their whitelist does not allow access to the Plex servers.</p><h3>The Code</h3><p>The <a href="https://github.com/Tyzer34/plexMusicPlayer"><em>PlexMusicPlayer</em></a> project is provided on GitHub and will be updated in the future. The purpose of this project is to give Alexa users a way to playback music, using their own Plex server. However, it is also a way to show developers a method to integrate Plex into Alexa as a music service. The application does however not enable remote control of Plex instances, nor does it (as for this moment) include a global technique of authentication. However, a detailed overview on how to setup this application is provided in the next section.</p><h3>The Setup</h3><p>As previously mentioned, I deployed this project using Heroku. Their free tier allows a single web application to be run upon a free <em>dyno.</em> However, keep in mind that this dyno is limited to 550 usage hours per month (or 1000 if verified with a credit card) and that the application goes into sleep mode after 30 minutes of receiving no web traffic. In their FAQ, they say the following about the usage of a free dyno:</p><blockquote>If an app has a web dyno that runs using a Free dyno, and that web dyno receives no traffic in a 30 minute period, the web dyno will sleep. In addition to the web dyno sleeping, the worker dyno (if present) will also sleep. Free dynos do not consume Free dyno hours while sleeping.</blockquote><p>This means that the free tier is more than enough for self-deployment of the application. However, keep in mind that if the application goes into sleep mode, it takes some time to start up again. This has a big impact on the application, as the first query by the user will result in a timeout, responding with the following error.</p><pre>The remote endpoint could not be called, or the response it returned was invalid.</pre><p>Even though this first query fails, it produces web traffic to our application deployed on Heroku, allowing it to wake up. After a small twenty seconds or so, the wake up sequence has finished and the application can be used.</p><h4>Deployment on Heroku</h4><p>First, you should have Heroku and Git installed on your local machine. If you have not yet worked with Heroku before, the installation process can be found <a href="https://devcenter.heroku.com/articles/heroku-cli">here</a>. Also, make sure that you are already logged in with Heroku account.</p><p>Next, we want to clone the current GitHub repository of this project. First, open up a terminal (or command prompt) and navigate to the directory you want the files to be stored in. Then, use the following command to clone the repo.</p><pre>git clone <a href="https://github.com/tyzer34/plexmusicplayer.git">https://github.com/tyzer34/plexmusicplayer.git</a></pre><p>Now that the files have been cloned locally, they can be deployed on Heroku. To set up this Heroku server, the following commands should be used. Note that <em>*yourServerName* </em>should be replaced by something else.</p><pre>heroku create *yourServerName*<br>git push heroku master</pre><p>The application should normally be online now! Now we just have to set the authentication variables on the server and set up the Alexa Skill.</p><h4>Getting Authentication Variables</h4><p>In order to authenticate the PlexMusicPlayer application, two environmental variables need to be set, namely the <strong>Plex token</strong> and the <strong>base url</strong>.</p><p>First, you will need to navigate to <a href="http://plex.tv/web/app">plex.tv/web/app</a>, select a random episode, show, etc. and click the three dots on the left (as shown in the picture below in Dutch). Next, right click the <em>Download</em> option and select <em>copy link address.</em></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/355/1*XMIJSwQg5V8Rh4E8W4oYGQ.png" /></figure><p>The copied link address form should normally look something like this (with the parts filled in though):</p><pre><a href="https://192-168-0-226.9c04bc5196e2483bbe7919e96a0e5c1a.plex.direct:32400/library/parts/45/1473368932/file.mkv?download=1&amp;X-Plex-Token=7JQRYQYbXS2ZbW4oH7rx">https://your-plex-ip-address.some_long_encoded_string.plex.direct:your_plex_port/library/parts/some_id/some_file_id/file.ext?download=1&amp;X-Plex-Token=y</a>our_plex_token</pre><p>When looking at this long url, both needed variables can be extracted as following. It is important for the base url that the / after <em>your_plex_port</em> is not copied with.</p><ul><li><strong>base url</strong> = <a href="https://192-168-0-226.9c04bc5196e2483bbe7919e96a0e5c1a.plex.direct:32400/library/parts/45/1473368932/file.mkv?download=1&amp;X-Plex-Token=7JQRYQYbXS2ZbW4oH7rx">https://your-plex-ip-address.some_long_encoded_string.plex.direct:your_plex_port</a></li><li><strong>Plex token</strong> = <a href="https://192-168-0-226.9c04bc5196e2483bbe7919e96a0e5c1a.plex.direct:32400/library/parts/45/1473368932/file.mkv?download=1&amp;X-Plex-Token=7JQRYQYbXS2ZbW4oH7rx">y</a>our_plex_token</li></ul><p>Now, we have to set these environmental variables in our Heroku server. To do this, go back to the terminal (or command prompt) and enter the following lines, changed with your extracted variables.</p><pre>heroku config:set PLEX_TOKEN=your_plex_token<br>heroku config:set PLEX_URL=<a href="https://192-168-0-226.9c04bc5196e2483bbe7919e96a0e5c1a.plex.direct:32400/library/parts/45/1473368932/file.mkv?download=1&amp;X-Plex-Token=7JQRYQYbXS2ZbW4oH7rx">https://your-plex-ip-address.some_long_encoded_string.plex.direct:your_plex_port</a></pre><p>Lastly, we just have to restart the server so it can cope with the set environmental variables. To do this, just use the following command.</p><pre>heroku restart</pre><h4>Making the Alexa Skill</h4><p>Now that the backend is fully up and running, the Alexa Skill has to be initialized on <a href="http://developer.amazon.com/">Amazon Developers</a>. If you have not done this before, you can follow this <a href="https://blog.craftworkz.co/flask-ask-a-tutorial-on-a-simple-and-easy-way-to-build-complex-alexa-skills-426a6b3ff8bc#0acf">step-by-step explenation</a>.</p><p>As for the invocation name, I chose <em>plexmusic. </em>However, this can be anything you want. The information needed to fill in the Intent Schema and Sample Utterances can be found in the <a href="https://github.com/Tyzer34/plexMusicPlayer/tree/master/speech_assets">GitHub repository</a>, in the <em>speech_assets </em>folder.</p><p>Next, the endpoint needs to be set to <em>HTTPS</em> with the following link (with the <em>*yourServerName* </em>changed to the previously selected one)</p><pre>https://*yourServerName*.herokuapp.com/plex</pre><p>Lastly, select the 2nd option at <em>certificate</em> for endpoint and save your application.</p><p>Congratulations! PlexMusicPlayer is now up and running on your favorite Alexa device(s). Now, you are ready to try out some of its functionalities!</p><h3>The Functionalities</h3><p>Currently, PlexMusicPlayer is able to playback music based on a provided artist, album or song title. When an artist or album is provided and multiple songs are available for that artist or album in your Plex library, these will be queued and played as a playlist. Furthermore, it is possible to ask information related to the current song or to shuffle the current playlist. Obviously, requests such as next and play are also implemented. Some activation sentences are provided as an example below. More can be found in the <a href="https://github.com/Tyzer34/plexMusicPlayer/blob/master/speech_assets/sampleUtterances.txt">sample utterances file</a>.</p><ul><li>Alexa, ask plexmusic to play Def Leppard.</li><li>Alexa, ask plexmusic to play the album AM by Arctic Monkeys.</li><li>Alexa, ask plexmusic to tell me the name of this song.</li><li>Alexa, ask plexmusic to play “my favorites” playlist.</li></ul><h3>The Future</h3><p>As this is only an initial implementation, there is a lot of room for improvement and extra functionality. Below is a list of some of the things I’d like to see implemented in the project. Please note that there is no distinction between minor and major improvements in these examples.</p><ul><li>Enabling user authentication based on a Plex PIN (<a href="https://github.com/overloadut/node-plex-api-pinauth">node.js implementation</a>)</li><li>Implementing more playback functionalities, such as playback based on genre or year</li><li>Adding Alexa cards to show the requested song, album or artist with a picture</li><li>Making more intents, so the user can request specific actions such as what year the song is from</li></ul><p>Tracking of these future implementations are provided in the GitHub repository. User contributions and improvements are also welcome!</p><p>Hopefully, this guide was able to allow music playback from your Plex Media Server on Echo devices. If there are any issues or questions, please do not hesitate to open up an issue on GitHub.</p><figure><a href="http://bit.ly/HackernoonFB"><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*0hqOaABQ7XGPT-OYNgiUBg.png" /></a></figure><figure><a href="https://goo.gl/k7XYbx"><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Vgw1jkA6hgnvwzTsfMlnpg.png" /></a></figure><figure><a href="https://goo.gl/4ofytp"><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*gKBpq1ruUi0FVK2UM_I4tQ.png" /></a></figure><blockquote><a href="http://bit.ly/Hackernoon">Hacker Noon</a> is how hackers start their afternoons. We’re a part of the <a href="http://bit.ly/atAMIatAMI">@AMI</a> family. We are now <a href="http://bit.ly/hackernoonsubmission">accepting submissions</a> and happy to <a href="/service/mailto:partners@amipublications.com">discuss advertising &amp; sponsorship</a> opportunities.</blockquote><blockquote>If you enjoyed this story, we recommend reading our <a href="http://bit.ly/hackernoonlatestt">latest tech stories</a> and <a href="https://hackernoon.com/trending">trending tech stories</a>. Until next time, don’t take the realities of the world for granted!</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*35tCjoPcvq6LbB3I6Wegqw.jpeg" /></figure><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fupscri.be%2Fdde502%3Fas_embed%3Dtrue&amp;dntp=1&amp;url=https%3A%2F%2Fupscri.be%2Fhackernoon%2F&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=upscri" width="800" height="400" frameborder="0" scrolling="no"><a href="https://medium.com/media/3c851dac986ab6dbb2d1aaa91205a8eb/href">https://medium.com/media/3c851dac986ab6dbb2d1aaa91205a8eb/href</a></iframe><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=38b14b41faf0" width="1" height="1" alt=""><hr><p><a href="https://medium.com/hackernoon/plex-alexa-the-perfect-wedding-38b14b41faf0">Plex ♥ Alexa — The perfect wedding</a> was originally published in <a href="https://medium.com/hackernoon">HackerNoon.com</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Flask-Ask — A tutorial on a simple and easy way to build complex Alexa Skills]]></title>
            <link>https://medium.com/craftworkz/flask-ask-a-tutorial-on-a-simple-and-easy-way-to-build-complex-alexa-skills-426a6b3ff8bc?source=rss-75cde01c7b82------2</link>
            <guid isPermaLink="false">https://medium.com/p/426a6b3ff8bc</guid>
            <category><![CDATA[python]]></category>
            <category><![CDATA[alexa]]></category>
            <category><![CDATA[flask]]></category>
            <category><![CDATA[amazon-echo]]></category>
            <category><![CDATA[chatbots]]></category>
            <dc:creator><![CDATA[Bjorn Vuylsteker]]></dc:creator>
            <pubDate>Mon, 26 Dec 2016 19:45:07 GMT</pubDate>
            <atom:updated>2016-12-26T19:45:07.552Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Zj9En-11PlMQITsoc2N3UA.jpeg" /></figure><p>The age of robotics is here, and thus, many new toys are emerging. I managed to get my hands on an <a href="https://www.amazon.com/All-New-Echo-Dot-2nd-Generation/dp/B01DFKC2SO">Echo Dot (2nd generation)</a>, which includes Amazon’s personal assistant Alexa, that allows voice-enabled control of your smart environment. However, it can do so much more than that! You could for example ask for general information like <em>“Alexa, when were the Arctic Monkeys founded?”</em>, play music right from your Spotify premium account or just do generic tasks like setting a timer or an alarm. Thankfully, Amazon has recently enabled <em>Alexa Skills</em>, which allow developers to build applications for deployment upon the Echo devices. As I like to experiment with the latest technology, I tried to find out the easiest and cheapest way to build an Alexa skill, so you can make your own at home! I have also provided a <a href="https://github.com/Tyzer34/Flask-Ask_ExampleApplications">GitHub repository</a>, where you can find the full implementation.</p><h3>The Requirements</h3><p>First of, I set some requirements for myself to make this challenge a bit more interesting.</p><ul><li>The Alexa skills should be easily deploy-able and manageable</li><li>They should be able to execute complex query’s if needed</li><li>As previously mentioned, everything should be provided for free</li></ul><p>Starting of this journey, I quickly found myself the <a href="https://flask-ask.readthedocs.io/en/latest/"><em>Flask-Ask</em></a> library. It is based upon the very popular Python <em>Flask</em> library, which allows developers to, for example, deploy REST microservices with ease. Flask-ask, however, provides the easy deployment of Alexa skills via a similar structure. As I am already a huge Flask / Python fanatic, this library provides the ultimate solution for me.</p><p>Next, I wanted to deploy this application as a fully active cloud service. However, few providers actually give up server space for free. Luckily, I stumbled upon <a href="https://www.pythonanywhere.com/"><em>PythonAnywhere</em></a>, which allows users to deploy a single Flask server, among many things. Even though this limits the amount of possible Flask-Ask services to… well… one, as you will see, this won’t form any problem with the deployment of multiple Alexa skills.</p><h3>The Challenge</h3><p>As previously mentioned, I wanted to deploy multiple applications. However, I also wanted these applications to be complex enough, to show the huge amount of interesting things you could do with simple implementations. Therefore, I decided to build two applications:</p><ol><li>The first application will allow me to get the latest reviews for the shows I love from the <a href="http://ign.com/tv/reviews">IGN website</a>. Here, on top of the basics, a multitude of advanced techniques is used such as extracting web data, making Alexa cards, and separating code and speech using <em>Jinja</em>,</li><li>The second application will be a called <em>“I’m going on a trip”.</em> This will include context awareness and will show that multiple Alexa skills can be deployed using a single <em>Flask-Ask</em> server.</li></ol><p>In order to deploy these applications, we will have to set up the Flask webserver. To do this with PythonAnywhere, you will have to register for an account. Next, in the <em>Web</em> tab, you will have to add a new web app. During configuration, you have to select <em>Flask</em> as Python Web framework. Next select <em>Python 2.7</em> as preferred Python version and (optionally) rename your file to <em>alexa.py</em>.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/613/1*tds3XojMfoufchVNNXjyyA.png" /></figure><p>Next, the <em>Flask-Ask</em> library will have to be installed, as this is not (yet) standard included for all PythonAnywhere users. First, go to the <em>Consoles </em>tab on PythonAnywhere and start a new <em>Bash</em> console. When booted up, simply run the following command:</p><pre>pip install —-user flask-ask</pre><p>Now, we are all set up to build the first application!</p><h3>The Magic</h3><p>In order to be able to build an Alexa Skill, we must first come to understand how Alexa handles its voice commands. Before all, it is important to note that the Amazon Echo Dot can actually be seen as a thin client, as all voice processing happens in cloud services.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Yia-9lmP9FBY7WbdKzPmRQ.png" /><figcaption>Communication setup with an Amazon Echo Dot, using a Python Flask-Ask server deployed upon PythonAnywhere</figcaption></figure><p>First, speech is send to the Alexa Voice Services (AVS), where the appropriate Alexa Skill requested is selected. If an intent (action) was provided during speech, this intent will be sent towards the selected Alexa Skill implementation, together with potential provided variables. If no intent was provided, e.g. opening the application, the basic <em>Launch</em> intent will be called upon.</p><p>Next, based on the provided intent, a certain method will be called upon in the deployed Flask-Ask service. This method will, for example, do a certain calculation based upon the provided variables and will formulate an appropriate response. This response is then send back to the Amazon Echo device, via the AVS. As seen on the figure above, two different responses are available. Logically, a textual response can be provided, which will be executed by the Echo device in the form of speech. However, a <em>card</em> can also be made, providing extra visual information to the users via the Alexa app.</p><h3>The first application — IGN Reviews</h3><p>Now that we have established how the Amazon Echo devices operate, we can start building our first application! In order to make sure the developed Alexa skill is available to our Echo device, both the skill and AVS need to be implemented, working together to form the desired application. First, the back-end logic will be implemented. Afterwards, the AVS will be set-up to deploy our skill onto Echo device.</p><h4>The Alexa Skill —Providing Intelligence</h4><p>To start implementation of this first application, we will have to navigate to the <em>alexa.py</em> file in PythonAnywhere. To do this, click the <em>Files</em> tab and next, click on <em>mysite/</em> under Directories. Here, click on <em>alexa.py </em>to open up our webserver implementation. You will see, that as an example, a very simple “Hello World” app is provided. As we want to start from scratch, the provided code can be wiped.</p><p>First of all, we will import the needed libraries used in this first application. <em>Flask</em> and <em>Flask-Ask</em> are used to set up the Alexa skill, <em>datetime</em> will be used to sort reviews by time, <em>requests</em> will be used to access external services, <em>urllib</em> will be used to encode url strings and <em>re</em> will be used for filtering needed data.</p><pre>from flask import Flask, render_template<br>from flask_ask import Ask, statement, question, session<br>from datetime import datetime<br>import requests<br>import urllib<br>import re</pre><p>Next, the Alexa skill needs to be initialized and a first welcome message should be provided. The initialization of this endpoint is set under the <em>/alexa</em> subdomain. The welcome message is provided by implementing a function underneath the basic <em>launch</em> intent.</p><pre>app = Flask(__name__)<br>ask = Ask(app, &quot;/alexa&quot;)</pre><pre><a href="http://twitter.com/ask">@ask</a>.launch<br>def new_ask():<br>    welcome = render_template(&#39;welcome&#39;)<br>    reprompt = render_template(&#39;reprompt&#39;)<br>    return question(welcome) \<br>        .reprompt(reprompt)</pre><p>Here, the text used to welcome users (<em>welcome</em>) and to rephrase if no input is given (<em>reprompt</em>) is loaded from an external file, called <em>templates.yaml</em>. This file should be created on PythonAnywhere, in the same folder as the <em>alexa.py</em> file. In this file, textual input is given, such that code and text can be separated. For now, we just define the <em>welcome</em> and <em>reprompt</em> text.</p><pre>welcome: Welcome, to one of my homebrew applications. What would you like me to do?</pre><pre>reprompt: I didn&#39;t quite get that. Please state again what you would like me to do.</pre><p>Now that a first welcome message is provided, we can put some actual intelligence in the application. First, we will implement several methods to get and process data from the IGN website.</p><pre>def processShowName(show):<br>    showSynonymDict = {}<br>    showSynonymDict[&#39;Marvels Agents of Shield&#39;] = &quot;Marvel&#39;s Agents of S.H.I.E.L.D.&quot;<br>    showSynonymDict[&#39;Legends of Tomorrow&#39;] = &quot;DC&#39;s Legends of Tomorrow&quot;<br>    if show in showSynonymDict:<br>        return showSynonymDict[show]<br>    else:<br>        return show</pre><p>Later, when defining the Alexa Voice Service, we will set up multiple possibilities of shows that we can ask the reviews for. The <em>processShowName</em> method allows for this pronounced show as a string to be transformed into the actual name of the show. This will enable the correct results to be found by the <em>getReviews</em> method. In this method, an other platform called <a href="https://www.import.io/"><em>import.io</em></a> is used. This allows for an easy way to build APIs from structured websites.</p><pre>def getReviews(show):<br>    url = (&quot;<a href="http://www.ign.com/search?page=0&amp;count=10&amp;filter=articles&amp;type=article&amp;q=review">http://www.ign.com/search?page=0&amp;count=10&amp;filter=articles&amp;type=article&amp;q=review</a> &quot; + show).replace(&quot; &quot;, &quot;%20&quot;)<br>    url = urllib.quote(url, safe=&#39;&#39;)<br>    urlRest = &quot;<a href="https://api.import.io/store/connector/87e7f75c-3d4c-45cb-93e9-305a52237e63/_query?input=webpage/url">https://api.import.io/store/connector/87e7f75c-3d4c-45cb-93e9-305a52237e63/_query?input=webpage/url</a>:&quot; + url + &quot;&amp;&amp;_apikey=e581e9228fc34f139c3031d520f27af635f6dc576c932b84a8fd6cce4cc4b6a0a8499e8fcc53a7e308bf321c7a6e22998ec481eb1ab47742b062d01d62e174fc95343090d71c6dc63a4255886bf546e1&quot;<br>    data = requests.get(urlRest).json()<br>    return data</pre><p>By using the <em>getReviews</em> method, an JSON object is returned, filled with reviews of the requested show. To obtain this, first, the search results page url is formed. Next, it is encoded and added into the self-made <em>import.io</em> API. Finally, this API will return the JSON object, containing a multitude of reviews of the requested show, together with some extra useful information.</p><pre>def processDates(data):<br>    dateDict = {}<br>    for i in range(0, len(data[&quot;results&quot;])):<br>        regex = re.compile(&#39;(?&lt;=http:\/\/<a href="http://www.ign.com//articles//)(.*)(?=\/.*)&#39;,">www.ign.com\/articles\/)(.*)(?=\/.*)&#39;,</a> re.IGNORECASE)<br>        date_str = regex.findall(data[&quot;results&quot;][i][&quot;searchitem_link_2&quot;])[0].split(&quot;/&quot;)<br>        dateDict[i] = datetime(int(date_str[0]), int(date_str[1]), int(date_str[2]))<br>    return dateDict</pre><p>As we are requesting the latest review for that show, we can sort all reviews in the JSON object by date, returning a dictionary with their position in the JSON array as key and their respective date-stamp as value. This date-stamp is extracted using a <em>regex</em> on a certain part of the acquired data.</p><pre>def getLatestReview(show):<br>    data = getReviews(show)<br>    dateDict = processDates(data)<br>    i_latest = 0<br>    for i in dateDict:<br>        if dateDict[i] &gt; dateDict[i_latest]:<br>            i_latest = i<br>    latest = data[&quot;results&quot;][i_latest]<br>    return latest</pre><p>Lastly, we put it all together using the <em>getLatestReview</em> method. This will receive a certain show as a string, get the JSON data with reviews from that show, get the dictionary containing the date-stamps for those reviews and return the JSON data for the latest review.</p><p>Now, we can put this logic to actual use in the voice application. In order to provide the intended results, we need to define new speech templates in the <em>templates.yaml</em> file.</p><pre>latest_review: The latest episode of {{ show }}, called, {{ episode }}, got a score of {{ score }} on IGN.</pre><pre>reprompt_show: I didn&#39;t quite understand that. What show did you want the review from?</pre><p>To implement these speech templates, we create a response to the intent <em>ReviewLatestIntent</em>. This intent will be called upon when the AVS gets the appropriate question to give the latest review on a certain show.</p><pre><a href="http://twitter.com/ask">@ask</a>.intent(&#39;ReviewLatestIntent&#39;)<br>def launchReview(show):<br>    if (show is None):<br>        reprompt_show = render_template(&quot;reprompt_show&quot;)<br>        return question(reprompt_show)<br>    else:<br>        filShow = processShowName(show)<br>        latest = getLatestReview(filShow)<br>        review_title = latest[&quot;searchitem_link_2/_text&quot;]<br>        episode_title = review_title.split(&#39;:&#39;)[1].replace(&#39;&quot;&#39;,&#39;&#39;)<br>        if (&quot; Review&quot; in episode_title):<br>            episode_title = episode_title.replace(&quot; Review&quot;, &quot;&quot;)<br>        image_url = latest[&quot;searchitem_image&quot;].replace(&quot;_160w&quot;, &quot;_480w&quot;).replace(&quot;http&quot;, &quot;https&quot;)<br>        score = str(latest[&quot;reviewscore_number&quot;])<br>        description = latest[&quot;ignreview_description&quot;]<br>        latestReview_msg = render_template(&#39;latest_review&#39;, show=show, episode=episode_title, score=score)<br>        return statement(latestReview_msg) \<br>            .standard_card(title=review_title + &quot; - &quot; + score,<br>                            text=description,<br>                            small_image_url=image_url,<br>                            large_image_url=image_url)</pre><p>Note that the <em>launchReview </em>method has a variable called <em>show</em>. This will be provided by the Alexa Voice Service, and will either be a show from a range of possibilities or a <em>NoneType</em> object. Therefore, we first need to check whether the provided show is valid or not. If the show is not valid, we again ask for voice input of this show. If it is valid, we convert it to its proper name, get the latest review, and extract some data from the JSON object for clarity. Using this data, we then fill in the gaps in the <em>latest_review</em> speech template and return this as a voice response.</p><p>However, we also return a visual response, in the form of a standard card. This will allow the review to be shown on the Alexa app, accompanied with the appropriate picture and text. As for the simplicity of the application, a short description is given as text for the card. I looked for the possibility of adding a link to the review page, so it can be navigated to quickly. Having found <a href="https://forums.developer.amazon.com/questions/43706/using-urls-in-alexa-skill-cards.html">another person</a> asking this question, Amazon has stated the following:</p><blockquote>The Alexa Skill Kit does not support clickable links. Including a URL text string in a card is not recommended since that may present a poor and confusing user experience.</blockquote><p>Reading this statement, it does not seem like they even want to add url functionality into a card. However, using <em>import.io</em>, it would be possible to quickly get the full review in text and display this on the card, removing the need for the url.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/923/1*J7Mbq-OLu_Vj-l2gYzPvCw.png" /><figcaption>The card for the latest review of The Flash, as shown in the <a href="http://alexa.amazon.com/">Alexa web application</a></figcaption></figure><p>To make sure the Flask-Ask application is runnable, add the following lines of code to the <em>alexa.py</em> file. Finally, go to the <em>Web</em> tab in PythonAnywhere and reload the set up webserver.</p><pre>if __name__ == &#39;__main__&#39;:<br>    app.run(debug=True)</pre><p>Great! We have just deployed our first Alexa Skill! However, we have not yet made the link between the Echo device and our application. To do this, we have to create the communication part of the Alexa Skill.</p><h4>The Alexa Skill — Providing Communication</h4><p>First, you have to login with your Amazon credentials into their <a href="https://developer.amazon.com/home.html">developer console</a>. Having done this, navigate to the <em>Alexa</em> tab and select the <em>Alexa Skills Kit.</em> Click on <em>Add a New Skill</em> and we can start filling in the details.</p><p>As for <strong>Skill Information</strong>, we provide <em>IGN Reviews</em> as name for the application and <em>reviews</em> as invocation name. The choice of invocation name is really important, as this will be used to call our service with.</p><p>Next, the <strong>Interaction Model</strong> needs to be provided. First, we will create a list of possible shows, by adding a new custom slot type. For this slot, the type will be <em>LIST_OF_SHOWS </em>and the values are listed as following:</p><pre>Arrow<br>Legends of Tomorrow<br>The Flash<br>Supergirl<br>Marvels Agents of Shield<br>Westworld</pre><p>Now, the intent schema needs to be filled in. As we have only defined one intent into our application, we will only call upon this single intent. Notice that this intent contains a <em>slots</em> list. In the Alexa skill, this corresponds to the acquired input variables for our <em>launchReview</em> method.</p><pre>{<br>    &quot;intents&quot;: [<br>        {<br>           &quot;intent&quot;: &quot;ReviewLatestIntent&quot;,<br>           &quot;slots&quot;: [{<br>               &quot;name&quot;: &quot;show&quot;,<br>               &quot;type&quot;: &quot;LIST_OF_SHOWS&quot;<br>            }]<br>        }<br>    ]<br>}</pre><p>Lastly, we define sample utterances. These are example phrases that will correspond to the user interaction with the skill. Even though I have only defined four different possibilities, the Alexa skill will be able to cope with similar speech input.</p><pre>ReviewLatestIntent give me the latest review on {show}<br>ReviewLatestIntent give me the latest {show} review<br>ReviewLatestIntent what the latest {show} review is<br>ReviewLatestIntent how the last {show} episode scored</pre><p>To be able to communicate with our Flask-Ask application, we need to set up the <strong>Configuration</strong> correctly. As a service endpoint type, select <em>HTTPS</em>, your geographical preference and paste the url to your Flask-Ask service here. (<a href="https://tyzer34.pythonanywhere.com/alexa">https://*yourUserName*.pythonanywhere.com/alexa</a>)</p><p>As we are not using authenticated third party services, we can ignore the Account Linking and click <em>Next</em>. Lastly, to setup the <strong>SSL Certificate</strong>, select the second option and <em>Save</em> the application.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/811/1*sTLzwWVVfA-zocUe9vrkhg.png" /></figure><p>Congratulations! The homebrew Alexa skill is now fully deployed and operational! You can now test it out with your favorite Echo device or use the <a href="https://echosim.io/">Echosim.io</a> service to test it out directly on your computer.</p><iframe src="https://cdn.embedly.com/widgets/media.html?url=http%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3Dk5NZ8zT0sHo&amp;src=https%3A%2F%2Fwww.youtube.com%2Fembed%2Fk5NZ8zT0sHo&amp;type=text%2Fhtml&amp;key=d04bfffea46d4aeda930ec88cc64b87c&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/46279c025ad8d11298f63e8c975c68b0/href">https://medium.com/media/46279c025ad8d11298f63e8c975c68b0/href</a></iframe><h3>The second application —“I’m going on a trip”</h3><p>To prove that two Alexa skills can be deployed upon a single Flask-Ask service and to show context awareness possibilities in app development, we are going to create a second application called “I’m going on a trip”. As proof of concept, the application will provide us with a certain European city. Next, we can ask follow up questions about that city, like the weather forecast. An example scenario of how an interaction could go, is provided below.</p><ul><li>Alexa] They say London is beautiful this time of the year.</li><li>User] What’s the weather like over there?</li><li>Alexa] In London, the weather is currently being described as cloudy, with a temperature of 2 degrees Celsius.</li><li>User] Give me another city.</li><li>Alexa] A great place to visit might be Venice.</li><li>User] How is the weather?</li><li>Alexa] In Venice, the weather is currently being described as sunny, with a temperature of 11 degrees Celsius.</li></ul><p>Following this example scenario, we will start the implementation of our second application.</p><h4>The Alexa Skill — Providing Intelligence</h4><p>First of all, we are going to define some new speech responses in the <em>templates.yaml</em> file.</p><pre>trip_open: Hi there! Are you ready to go on a trip with me?</pre><pre>trip_again: Are you still here? Tell me when you&#39;re ready!</pre><pre>trip_start: Okay great!</pre><pre>trip_city_1: They say {{ city }} is beautiful this time of the year.</pre><pre>trip_city_2: A great place to visit might be {{ city }}.</pre><pre>trip_city_3: Have you ever been to {{ city }}? It&#39;s lovely!</pre><pre>trip_city_propose_good: Oh yeah! I haven&#39;t been to {{ city }} yet. Great idea.</pre><pre>trip_city_propose_bad: Hmm. I&#39;m not really feeling like going there. Do you have another proposal?</pre><pre>trip_weather: In {{ city }}, the weather is currently being described as cloudy, with a temperature of {{ temp }} degrees Celsius.</pre><pre>trip_go: Great! I&#39;ve already booked us a flight for tomorrow! Let&#39;s start packing!</pre><pre>trip_nogo: But I really wanted to go on a trip with you.</pre><p>Next, we can begin the implementation of the application. To do this, we open up the <em>alexa.py</em> file, where our Alexa skill is implemented. Here, we define an intent called <em>TripWelcomeEvent</em>, which will be called upon when asking the application to go on a trip. The implementation of this intent is fairly straightforward.</p><pre><a href="http://twitter.com/ask">@ask</a>.intent(&#39;TripWelcomeIntent&#39;)<br>def trip_welcome():<br>    welcome = render_template(&#39;trip_open&#39;)<br>    reprompt = render_template(&#39;trip_again&#39;)<br>    return question(welcome) \<br>        .reprompt(reprompt)</pre><p>When the user is being asked if he or she is ready, a positive and negative response can be provided. Instead of making our own intents to handle this, we can use standard Amazon intents that have already been extensively tested. First, we will implement the <em>trip_nogo </em>template. This should be called upon when the user isn’t ‘ready’ or when the application is cancelled or stopped. As this comprises multiple intents for the same method, an implementation is provided as follows.</p><pre><a href="http://twitter.com/ask">@ask</a>.intent(&#39;AMAZON.CancelIntent&#39;)<br><a href="http://twitter.com/ask">@ask</a>.intent(&#39;AMAZON.StopIntent&#39;)<br><a href="http://twitter.com/ask">@ask</a>.intent(&#39;AMAZON.NoIntent&#39;)<br>def trip_nogo():<br>    quit = render_template(&#39;trip_nogo&#39;)<br>    return statement(quit)</pre><p>Before adding the <em>YesIntent</em>, we will first have to implement the basic logic of the application. As the purpose of the application is to suggest a trip to a certain European city, we want a list of cities from which we can select. This is provided by the <em>getCityList</em> method. For simplicity, we are manually going to put together this list of cities. However, an external service or database could also be used.</p><pre>def getCityList():<br>    list = [&#39;Paris&#39;, &#39;London&#39;, &#39;Rome&#39;, &#39;Madrid&#39;, &#39;Prague&#39;, &#39;Venice&#39;, &#39;Florence&#39;, &#39;Dublin&#39;, &#39;Kopenhagen&#39;]<br>    return list</pre><p>We also need a method that can return a randomly selected element from the provided list. To use the following method, we need to import the <em>random</em> library.</p><pre>import random</pre><pre>def getRandomElement(list):<br>    return random.choice(list)</pre><p>To add some randomness to the provided speech outputs, we also implement a <em>getRandomCityText</em> method.</p><pre>def getRandomCityText(city):<br>    texts = [&#39;trip_city_1&#39;, &#39;trip_city_2&#39;, &#39;trip_city_3&#39;]<br>    choice = getRandomElement(texts)<br>    cityText = render_template(choice, city=city)<br>    return cityText</pre><p>Now, we are going to summarize the application logic. To do this, a form of memory storage will be set up, which will be available only during the duration of the session. This implies that, if the application is cancelled, stopped or has ended, the stored variables are removed. To implement the standard <em>YesIntent</em> from Amazon, first, select a random city from the provided list of cities. Next, in our session attributes, we make a <em>currentCity</em> and <em>promptedCities</em> variable. This last variable will contain a list with cities already proposed during this session. Lastly, we return speech as a question. This is very important, as a question response does not close the current session, where a statement closes it, losing the saved variables.</p><pre><a href="http://twitter.com/ask">@ask</a>.intent(&#39;AMAZON.YesIntent&#39;)<br>def trip_start():<br>    start = render_template(&#39;trip_start&#39;)<br>    cityList = getCityList()<br>    city = getRandomElement(cityList)<br>    promptedCities = [city]<br>    session.attributes[&#39;promptedCities&#39;] = promptedCities<br>    session.attributes[&#39;currentCity&#39;] = city<br>    cityPhrase = getRandomCityText(city)<br>    return question(start + cityPhrase)</pre><p>Next, we will implement the <em>TripNextCityIntent</em>, which will be called upon when the user wants another suggestion. The main difference here, is that a check is applied to make sure that a certain city is not prompted twice during a session.</p><pre><a href="http://twitter.com/ask">@ask</a>.intent(&#39;TripNextCityIntent&#39;)<br>def trip_nextCity():<br>    cityList = getCityList()<br>    city = getRandomElement(cityList)<br>    promptedCities = session.attributes[&#39;promptedCities&#39;]<br>    while city in promptedCities:<br>        # This can go into a loop if all cities have been prompted<br>        city = getRandomElement(cityList)<br>    promptedCities.append(city)<br>    session.attributes[&#39;promptedCities&#39;] = promptedCities<br>    session.attributes[&#39;currentCity&#39;] = city<br>    cityPhrase = getRandomCityText(city)<br>    return question(cityPhrase)</pre><p>In order to let the user have a certain degree of freedom, we will allow the proposal of a city to go to. This will be implemented by the <em>TripProposeIntent</em>, where a city will be provided to the method.</p><pre><a href="http://twitter.com/ask">@ask</a>.intent(&#39;TripProposeIntent&#39;)<br>def trip_propose(city):<br>    if (city is None):<br>        response = render_template(&#39;trip_city_propose_bad&#39;)<br>        return question(response)<br>    else:<br>        response = render_template(&#39;trip_city_propose_good&#39;, city=city)<br>        promptedCities = session.attributes[&#39;promptedCities&#39;]<br>        promptedCities.append(city)<br>        session.attributes[&#39;promptedCities&#39;] = promptedCities<br>        session.attributes[&#39;currentCity&#39;] = city<br>        return question(response)</pre><p>Now that we can ask for suggestions and input our own, it is time to make a follow up question. For simplicity of this application, we are only going to make a single one, namely asking about the weather. For the data, we will use an API provided by <a href="https://openweathermap.org/">OpenWeatherMap</a>, which allows up to search for weather data of a city based on a string.</p><pre><a href="http://twitter.com/ask">@ask</a>.intent(&#39;TripWeatherIntent&#39;)<br>def trip_weather():<br>    city = session.attributes[&#39;currentCity&#39;]<br>    url = &#39;<a href="http://api.openweathermap.org/data/2.5/weather?q=&#39;">http://api.openweathermap.org/data/2.5/weather?q=&#39;</a> + city<br>    key = &#39;&amp;APPID=3d02070a84a923fe26dd362ac34ce327&amp;units=metric&amp;lang=en&#39;<br>    response = requests.get(url + key)<br>    descr = str(response.json()[&#39;weather&#39;][0][&#39;description&#39;])<br>    temp = str(int(response.json()[&#39;main&#39;][&#39;temp&#39;]))<br>    response = render_template(&#39;trip_weather&#39;, city=city, description=descr, temp=temp)<br>    return question(response)</pre><p>Lastly, we have to be able to plan the trip to the wanted city, so we implement the <em>TripGoIntent</em>. This will finalize the session and provide a card on the Alexa app that will show when your trip is planned.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/574/1*an1AF8Yj5KZttfyfVMmAXA.png" /><figcaption>The simple card response, as displayed on the Alexa Web Application.</figcaption></figure><pre><a href="http://twitter.com/ask">@ask</a>.intent(&#39;TripGoIntent&#39;)<br>def trip_go():<br>    city = session.attributes[&#39;currentCity&#39;]<br>    response = render_template(&#39;trip_go&#39;)<br>    return statement(reponse) \<br>        .simple_card(title=&#39;Your trip for &#39; + city, content=&quot;Don&#39;t forget to catch your plane tomorrow!&quot;)</pre><p>All right! The second application is all set up! Don’t forget to reload your web application. Now we just have to set up the communication part again.</p><h4>The Alexa Skill — Providing Communication</h4><p>Even though this second application is deployed on the same Flask-Ask service, we will create a new Alexa Skill in the Amazon Developer Console.</p><p>For <strong>Skill Information</strong>, I provided <em>I’m going on a trip</em> as the name and <em>trips</em> as the invocation name, in order to easily call out the application.</p><p>As we are using many intents, the <strong>Interaction Model<em> </em></strong>needs quite a bit of information. First, we declare the Intent Schema as follows.</p><pre>{<br>    &quot;intents&quot;: [<br>        {<br>            &quot;intent&quot;: &quot;TripWelcomeIntent&quot;<br>        },<br>        {<br>           &quot;intent&quot;: &quot;AMAZON.YesIntent&quot;<br>        },<br>        {<br>           &quot;intent&quot;: &quot;AMAZON.NoIntent&quot;<br>        },<br>        {<br>           &quot;intent&quot;: &quot;AMAZON.CancelIntent&quot;<br>        },<br>        {<br>           &quot;intent&quot;: &quot;AMAZON.StopIntent&quot;<br>        },<br>        {<br>           &quot;intent&quot;: &quot;TripProposeIntent&quot;,<br>           &quot;slots&quot;: [{<br>                &quot;name&quot;: &quot;city&quot;,<br>               &quot;type&quot;: &quot;AMAZON.EUROPE_CITY&quot;<br>            }]<br>        },<br>        {<br>           &quot;intent&quot;: &quot;TripNextCityIntent&quot;<br>        },<br>        {<br>           &quot;intent&quot;: &quot;TripWeatherIntent&quot;<br>        },<br>        {<br>           &quot;intent&quot;: &quot;TripGoIntent&quot;<br>        }<br>    ]<br>}</pre><p>Next, we set up multiple Sample Utterances, so the application is able to understand various ways of speech provided by the user. Note that, like for example <em>Thats</em>, not everything can be written according to the correct spelling. This is because Sample Utterances do not allow certain symbols to be included in the example phrases.</p><pre>TripWelcomeIntent plan a new trip<br>TripWelcomeIntent set up a trip<br>TripWelcomeIntent arrange a vacation</pre><pre>AMAZON.YesIntent I am ready<br>AMAZON.YesIntent Yes I am</pre><pre>TripProposeIntent I want to go to {city}<br>TripProposeIntent How about {city}<br>TripProposeIntent {city} seems nice<br>TripProposeIntent Can we go to {city}</pre><pre>TripNextCityIntent Give me another<br>TripNextCityIntent I want another city<br>TripNextCityIntent I do not want to go there<br>TripNextCityIntent next city</pre><pre>TripWeatherIntent What is the weather like over there<br>TripWeatherIntent How is the weather</pre><pre>TripGoIntent Seems great<br>TripGoIntent Lets go<br>TripGoIntent Lets go there<br>TripGoIntent That is good for me</pre><p>Lastly, both the <strong>Configuration</strong> and <strong>SSL Certificate</strong> are set up in exactly the same way as described in application one.</p><p>Hit that <em>Save</em> button and the second application is all set up! You can now test it out on your favorite Echo device!</p><iframe src="https://cdn.embedly.com/widgets/media.html?url=http%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DQUrbMjNbX_A&amp;src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FQUrbMjNbX_A&amp;type=text%2Fhtml&amp;key=d04bfffea46d4aeda930ec88cc64b87c&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/3f149a93ae820c400a2de9495df9635a/href">https://medium.com/media/3f149a93ae820c400a2de9495df9635a/href</a></iframe><h3>The Conclusion</h3><p>It was really fun toying around with the Amazon Echo Dot. Looking back at my initial requirements, this project has been a great success.</p><p>First of all, the Alexa skills are very easily deploy-able, as you only have to import the Flask-Ask library and add a single line in front of the method you want to set up as a service.</p><p>Next, the level of complexity can be as high as you want. Because we are using Python to deploy the Alexa skills with, the possibilities are endless. Nevertheless, delay might be an issue when the computational time exceeds the expected waiting time.</p><p>Lastly, everything presented in this tutorial is offered for free and can be used to deploy your own homebrew Alexa skills with.</p><p>I am already looking forward to building more Alexa skills and seeing new functionalities being added to both Alexa and the Flask-Ask library.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=426a6b3ff8bc" width="1" height="1" alt=""><hr><p><a href="https://medium.com/craftworkz/flask-ask-a-tutorial-on-a-simple-and-easy-way-to-build-complex-alexa-skills-426a6b3ff8bc">Flask-Ask — A tutorial on a simple and easy way to build complex Alexa Skills</a> was originally published in <a href="https://medium.com/craftworkz">Craftworkz</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Bye Pepper — See you again soon!]]></title>
            <link>https://medium.com/craftworkz/bye-pepper-see-you-again-soon-117359e9a99?source=rss-75cde01c7b82------2</link>
            <guid isPermaLink="false">https://medium.com/p/117359e9a99</guid>
            <category><![CDATA[robotics]]></category>
            <category><![CDATA[technology]]></category>
            <category><![CDATA[fun]]></category>
            <category><![CDATA[chatbots]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <dc:creator><![CDATA[Bjorn Vuylsteker]]></dc:creator>
            <pubDate>Fri, 16 Sep 2016 11:56:44 GMT</pubDate>
            <atom:updated>2016-09-16T11:56:44.705Z</atom:updated>
            <content:encoded><![CDATA[<p>During these past few weeks, I’ve had the opportunity to work on this amazing innovative Pepper project for Craftworkz. As I’m still an engineering student, getting this chance to work out a project full-scope means a lot to me. It was a working experience that I feel made me grow in knowledge and as a person. The colleagues here are amazing to work with and very supportive.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.facebook.com%2Fvideo%2Fembed%3Fvideo_id%3D1099439596758483&amp;url=https%3A%2F%2Fwww.facebook.com%2Fcraftworkzco%2Fvideos%2F1099439596758483%2F&amp;image=https%3A%2F%2Fscontent.xx.fbcdn.net%2Fv%2Ft15.0-10%2Fp128x128%2F14299416_1099441993424910_320671394_n.jpg%3Foh%3D6232ff830a82aff5a867385c6bf91b98%26oe%3D5841C7E0&amp;key=d04bfffea46d4aeda930ec88cc64b87c&amp;type=text%2Fhtml&amp;schema=facebook" width="560" height="315" frameborder="0" scrolling="no"><a href="https://medium.com/media/e07a202eabcc5b499631b924cb053ede/href">https://medium.com/media/e07a202eabcc5b499631b924cb053ede/href</a></iframe><p>I also had a fun time joking around with Pepper. I really feel like we bonded over those couple of weeks I got to work here. Sometimes, we even went out in the weekends. We had a fun time on the beach and our tour through Italy was so beautiful! I will have a hard time getting used to the fact that I won’t be seeing him again soon..</p><blockquote>“Never say goodbye because goodbye means going away and going away means forgetting.” ― <strong>J.M. Barrie</strong>, <strong>Peter Pan</strong></blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*jN7QQYSdHp_m6KNyJiICEQ.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*G1EPkycDSIL-71aQ3Fp82Q.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/796/1*i7MmepSWO38k29sAj4vrow.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/796/1*QWe2OgmM8RFDO0EiIMk61Q.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/796/1*JIVvt-wDWY-T6quO8eBbVQ.jpeg" /></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=117359e9a99" width="1" height="1" alt=""><hr><p><a href="https://medium.com/craftworkz/bye-pepper-see-you-again-soon-117359e9a99">Bye Pepper — See you again soon!</a> was originally published in <a href="https://medium.com/craftworkz">Craftworkz</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Meet Pepper - Sci-Fi or Reality?]]></title>
            <link>https://medium.com/craftworkz/meet-pepper-sci-fi-or-reality-735016f5c72f?source=rss-75cde01c7b82------2</link>
            <guid isPermaLink="false">https://medium.com/p/735016f5c72f</guid>
            <category><![CDATA[robotics]]></category>
            <category><![CDATA[prototyping]]></category>
            <category><![CDATA[pepper]]></category>
            <category><![CDATA[chatbots]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <dc:creator><![CDATA[Bjorn Vuylsteker]]></dc:creator>
            <pubDate>Wed, 07 Sep 2016 09:45:40 GMT</pubDate>
            <atom:updated>2016-09-08T15:17:52.600Z</atom:updated>
            <content:encoded><![CDATA[<p><em>Robot</em>, it is a term mentioned as far back as the 1920’s. The fictional tension of having a machine capable of human thinking, inspired many writers during the past century to come up with incredible science fictional stories. As time passed however, technology evolved beyond our wildest dreams and little by little, sci-fi became reality. The question is however, what has already been accomplished with our current technology and what remains science fiction?</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/392/1*bojf26j81aJlKWAVQZdwgQ.jpeg" /></figure><p>In the context of this question, we present you with the <strong>Pepper robot</strong>. It is described as a humanoid robot, capable of executing and comprehending human speech, while interacting with the user in different ways. It was build by Aldebaran as the intelligent brother for <em>Nao </em>and <em>Romeo</em>, with its main advantage being the unique capability to recognize principal human emotions. This way, Pepper can analyse how the user feels and respond accordingly.</p><p>For a prototyping project, we were asked to showcase what the Pepper is capable of in an unique and funny way. By analyzing the capabilities of Pepper, we came up with a personality analysis system, fully hidden from the users perspective.</p><h3>The brains behind the machine</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/787/1*DsvBzQGAYRzP2Vae-vpIBw.png" /><figcaption>Main Choreographe outlay for the Pepper software</figcaption></figure><p>As intelligent as Pepper seems to be, it is nothing without a little software to keep it up and running. Thankfully, Aldebaran made a software package called Choreographe to build just that! It lets you use a library of prebuilt codeblocks for use in fast prototyping. All codeblocks are Python-based, so they are fairly easy to understand and customize.</p><p>The software build for our Pepper prototype works as follows. On initial start up, some minor configuration will be done. Afterwards, Pepper will start with autonomous walking until a person is detected. When this happens, the robot will start his analysis of the person in front of it. It can deduce the persons age, gender, mood, smile, etc. and log this data to the cloud. When both the age and gender parameters have been set, a user interaction scenario will be selected accordingly. This consists of funny stories and questions, adjusted to the person in front. The answers to these questions will also be logged for further analysis. After the interaction is over, or after a timeout with no user response, Pepper will reset itself and reenter his autonomous walking state, and will start looking for the next user to interact with.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*MqAndt-1hCDoMjY6qKxTHA.png" /><figcaption>Basic schematic showing the Prototype workflow</figcaption></figure><h3>Robot of the future? - An Analysis</h3><p>Pepper itself is an amazing machine, with a huge amount of functionalities build into, but it is not that intelligent on its own. Thankfully, we live in an era where cloud technology is up and running. For our prototype for example, we used a Python Flask back-end API to store and get data from. Conversational interactions were programmed into Pepper by use of the <em>QiChat</em> dialogue system, but this could also be linked to an external chatbot. The possibilities of combining Pepper with cloud technology are endless, but as with any product, there are some limitations and restrictions to what can be done. To give a little insight on Pepper, we summed up a few pros and cons from our own experience with Pepper.</p><h4>PROS</h4><ul><li>Pepper is based on the<a href="http://doc.aldebaran.com/2-4/naoqi/index.html"> <em>NAOqi</em> framework</a>, which has an enormous amount of API’s available with Python and C++ support.</li><li>With the power of cloud technology, an intelligent AI framework can be build without having to worry about hardware issues.</li><li>Choreographe software suite lets you quickly test and prototype certain features.</li><li>Built-in animation libraries and contextual animation.</li><li>Emotional, facial and ambiance analysis features.</li><li>QiChat dialogue system, wich makes a great feature for voice interaction.</li></ul><h4>CONS</h4><ul><li>Pepper’s hardware setup doesn’t allow a huge amount of data processing.</li><li>Language support is currently limited to English and Japanese.</li><li>Because Choreographe software suite is block-based, big programs can become quite complex and overview of the current state of code decreases.</li><li>Dialogue system is almost unusable when <em>real</em> user interaction is required.</li></ul><h3>Ex Machina?</h3><p>People would say that humanoid AI robots are still a concept of science fiction, but are they? It is all in the way how we define a robot. The border between science fiction and reality is starting to fade and more and more <em>robots </em>are coming into households and public places as intelligent vacuum cleaners or voice enabled phones. Without realizing, we are already living in a century with robots around us. The concept of a robot being able to reflect full human interaction however, isn’t here (yet). Pepper is a great step towards enabling fully autonomous humanoid robots, but it lacks the proper intelligence. I’m certain however, that with the combination of cloud technology, Pepper could in fact be set up as an intelligent being. Research in AI has already reached a point of human understanding, context comprehension and adapted feedback. It’s just a matter of time before this kind of technology slips into the mind of a robot.</p><p>The next question however, is if (<em>when?</em>) robots will become self-aware, if they would be able to understand and execute sarcastic behavior for example and if they would be able to learn and deduce certain aspects of life like we humans do. At that point, I only can guess what comes next…</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*2FaxFChT4fJTbd3Ehxgm5g.jpeg" /></figure><blockquote>One day the AIs are going to look back on us the same way we look at fossil skeletons on the plains of Africa. An upright ape living in dust with crude language and tools, all set for extinction. — <em>Ex Machina</em></blockquote><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=735016f5c72f" width="1" height="1" alt=""><hr><p><a href="https://medium.com/craftworkz/meet-pepper-sci-fi-or-reality-735016f5c72f">Meet Pepper - Sci-Fi or Reality?</a> was originally published in <a href="https://medium.com/craftworkz">Craftworkz</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The Other Side… What’s it like?]]></title>
            <link>https://medium.com/@Tyzer34/whats-it-like-on-the-other-side-df0d185101f?source=rss-75cde01c7b82------2</link>
            <guid isPermaLink="false">https://medium.com/p/df0d185101f</guid>
            <dc:creator><![CDATA[Bjorn Vuylsteker]]></dc:creator>
            <pubDate>Mon, 11 Aug 2014 10:33:44 GMT</pubDate>
            <atom:updated>2014-09-18T07:09:40.676Z</atom:updated>
            <content:encoded><![CDATA[<p><em>A three-week students trip to China with Huawei’s Telecom Seeds for the future</em></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*gBRjo4Z4FXxoWV9WJq7S8w.jpeg" /><figcaption>5 august 2014<br>Getting a briefing at the Huawei headquarters in Brussels</figcaption></figure><h3>Leaving so soon?!</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*40cnaGk7ozlt8rhSLH_icw.jpeg" /></figure><p><em>[11/08]</em> “China, the country of innovation!” That’s what I&#39;ve been told throughout the history lessons a few years back. The question for me though, is if they can still be called the former? That, and many others are lurking around the corner, impatiently waiting to be answered. Fortunately, I can lay them to rest for a while, as they know for certain they’ll get their answers soon enough. Just in less than two weeks, I’m leaving for Beijing with the Huawei <em>Telecom Seeds for the Future </em>project. It’s a concept where Huawei sends a few selected students from an University to China, so they can learn about the culture and habits. It’s also a way for them to educate the next generation of students and to give them an (perhaps) experience that will stay with them forever. As you all have guessed by now, I am one of the lucky students to be selected for this program. For this, I had to go through a three round selection, involving an essay, a video and a face-to-face interview at Huawei Brussels. During the selections, I never hoped to get this opportunity, but you know what they say: “No guts, no glory!”.</p><p>As the departure date is coming closer, I thought of sharing my thoughts and viewpoints before I left. To be honest, I really don’t know what to expect of China itself. I’m not the kind of person to base thoughts on the prejudges of people, so I’m just going blank and open minded. As for the trip, I am expecting a lot! First of all, we’re gonna study a few days at the university of Beijing. This means that we’ll be able to see the cultural difference for students first handed! We’ll also be able so see some Chinese landmarks like the Olympic Stadium, The Chinese Wall, … and we’ll get education lessons in telecommunication by Huawei itself in Shenzhen! I’m awaiting many breathtaking moments, thoughtful idea’s and most of all an experience that I will be able to tell my grandchildren with great enthusiasm after all those years. Hopefully, you’ll follow me along on my three week journey, as I’ll get to experience China in a very unique and interesting way!</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*9L3d6633m5hlDOmixfoV8Q.jpeg" /><figcaption>Flight HU492<br>Brussels to Beijing<br>23 august<br>Departure: 13h45<br>(Belgian time)<br>Arrival: 5h30<br>(Chinese time)</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*e_86bNjHbS4ITEMkItd0Zg.jpeg" /><figcaption>Our Huawei correspondent for the first week, named Maria. She’s quite the energetic and caring person!</figcaption></figure><h3>Landed safely!!</h3><p><em>[24/08]</em> Yeah!! We’re in China!! The trip was long, but we made it safe and sound. It took us about 10 hours to fly from Brussels to Beijing. I was a bit worried that I wouldn&#39;t be able to keep me occupied during that time, but it passed swiftly by. The seats were comfortable (within reasonable marges) and the interactive tv-system was the perfect way to keep me busy. I watched some episodes of How I Met Your Mother, Friends, Top Gear, … and listened to a lot of music! There was also the option to learn a bit Chinese, to play some minigames, to watch movies, etc. It was a flight like I had never experienced before. I tried to catch some sleep, but it didn&#39;t work out so well. I’m not tired though at the moment (8h50 Chinese local time), so I’ll try to get a good nights rest this evening! As soon as we landed, we got our luggage really fast and followed our corresponding Huawei-colleague (in the picture) to the minivan. It was only a good half hour drive to our hotel, where we checked in quite quickly. We’re getting the chance to get some rest in our hotel room now, but our busy schedule will continue in a few short hours… I’m really excited about what we’re about to see today and the following three weeks!! Can’t wait!!</p><h3>A first impression..</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*zVRFz82uT6woFEuCrAJDCg.jpeg" /><figcaption>The first meal we had together at the restaurant of the Beijing Language University campus.</figcaption></figure><p><em>[24/08]</em> Wow, it’s already evening.. I’m not even aware of the fact that another day has passed this quickly! I guess the non-sleeping part could had had some influence in this.. As we decided not to sleep this morning — because we were not that tired — my roommate Klaas and me decided to go on a little walk. We ended up in a local supermarket where we bought a 5l water bottle. The tap water here isn&#39;t trustworthy, so we thought it would be best if we could brush our teeth with clean, filtered water. Not before long, the noon came and the first impression of a typical Chinese meal became a fact. At first, we were all led into a small room, as other people were sitting in groups in other different, separated rooms. We had to sit together at this big, round table with in the middle, a rotating glass-plate (see the picture above). The custom over here is simplified described as following. You count the number of people that you are eating together with and that will be the number of different dishes you will order. In that way, everyone will have the chance to taste many different things! This was a bit weird at first, because while there was still a lot of food left on the glass plate, more and more just kept on coming. I can’t say that I liked all the dishes, but for the most of them, I was surprised in a good way.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*le3q9zC6OK95CAMzEXomaA.jpeg" /><figcaption>The amazing, gold plated dragon figure at the pavilion in the garden of the Forbidden City</figcaption></figure><p>After lunch, we went on the bus to visit the Forbidden City. We saw the many different gates, leading up to the resident chambers of the emperors that once ruled this land. In his great garden, there was a pavilion with an amazing figure at the ceiling (picture left). Afterwards, we went on the little mountain to get a good view of the Beijing skyline and the Forbidden City (as viewed in the picture below). I was amazed by the symmetrical structure that had been put into this city. For the last activity of the day, we went to one of the best restaurants known for the local delicacy, the Peking Duck. Again, I was overwhelmed with the amount of food! For the Peking Duck itself, it was actually quite good! They prepared the meat up front, but we ate them wrapped up in a little thin pancake, with soy sauce and bits of melon. An odd combination, that surprisingly tasted well! After a little wait, we went back to the hotel, so we could get a great nights rest and that’s what I’m gonna do right away!</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*N9fTvlXEDrg6nZurBjADzw.jpeg" /><figcaption>A view from the rear entrance of the Forbidden City</figcaption></figure><h3>我叫比约恩！</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/720/1*0MpyK4f14ceyPe7UCjaflA.jpeg" /><figcaption>The openings ceremony, taking place in the Huawei courtyard, in the middle of Beijing</figcaption></figure><p><em>[26/08] </em>Two days have past since the last post, and I can say with certainty that this trip <em>will be </em>an experience that will stay with me for a lifetime! Yesterday, we had our first lesson in Mandarin. For the most part of the lesson, we were just pronouncing the different tones and accents, but we also learned some first words and how to sing Brother John (<em>Broeder Jacob</em>) and Happy Birthday in Chinese. This was very fitting as it was Klaas his birthday that day! After a nice, but little bit spicy lunch in the campus cafeteria, we had a lesson in calligraphy. For me personally, it was very hard as I’m not the kind of person to do very precise work with brush and ink. Still, I had fun trying my best to paint my given character. At the end of the class, we stood up and presented all our characters next to each other. The text eventually said: Happy Birthday Klaas! Afterwards, we got refreshed and suited up for the Opening Ceremony of the Huawei Telecom Seeds program. It was staged in an amazingly beautiful courtyard, in the middle of Beijing. As the weather was really clear, it was a beautiful evening. We had some drinks and speeches, there was live traditional Chinese music playing, we got ourselves a Chinese zodiac animal made out of clay, the food was really good (a mix of Belgian and Chinese food) and Klaas got an immense birthday cake. We had lots of fun that night and I really felt that the bond between us <em>seeds</em> was getting stronger and stronger.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*yJh_lqNkhxAAIQ0NV-xoRw.jpeg" /><figcaption>Francis and me practicing our Chinese, by holding a dialogue in front of the classroom</figcaption></figure><p>As the sun rose in the east, we awoke early in the morning, eager to learn more about this challenging language. We learned a few basic conversations and a bit of grammar. In the afternoon, we took our brush up once again. This time, it was to paint — another skill that I haven’t been able to master — and <a href="http://instagram.com/p/sJyr1bR_UY">the result</a> I got out of it looked really pretty (to me at least).</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*AVeCzhdGP5BFYU2jELrX0Q.jpeg" /><figcaption>Korean Hot Pot with a typical Korean drink (grapefruit juice)</figcaption></figure><p>After the class was over, we visited a well known Korean restaurant, that was located around the hotel area. The food there was so spicy, that almost everyone had to order another extra drink! Still, I really think that the degree of spiciness around here could be a lot worse. Back in the hotel, we picked up our textbooks and started learning in our rooms. As we have an exam planned on Friday, it’s really necessary that we practice this extraordinary language. It’s a lot to take in though and we’re moving along pretty fast, but practice makes perfect! After everyone was done, we went for a little trip to the city center. We took the metro and visited the local market. There we saw living scorpions, starfishes on a stick, octopus parts with soup, crickets, dried snakes, … The smell was sometimes unbearable, as if someone had just puked on the pavement. Later in the evening, we returned safe and sound to the hotel with the taxi. And so, this Telecom Seeds Program is really starting to become a reality for us, as we explore different bits of the culture and language in China.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=http%3A%2F%2Fscontent-b.cdninstagram.com%2Fhphotos-xaf1%2Ft50.2886-16%2F10648510_255091908019166_462865694_n.mp4&amp;src_secure=1&amp;url=http%3A%2F%2Finstagram.com%2Fp%2FsKubuCx_eD&amp;image=http%3A%2F%2Fphotos-e.ak.instagram.com%2Fhphotos-ak-xfa1%2F10654862_1470355636554708_1926445744_n.jpg&amp;key=d04bfffea46d4aeda930ec88cc64b87c&amp;type=video%2Fmp4&amp;schema=instagram" width="616" height="716" frameborder="0" scrolling="no"><a href="https://medium.com/media/62a84343f29ec901a3677256df7b95ea/href">https://medium.com/media/62a84343f29ec901a3677256df7b95ea/href</a></iframe><h3>Aced the exam!</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/640/1*3IaaEKWNE52NnO-RiPPkoA.jpeg" /><figcaption>Me holding my certificate that I passed the courses in the Beijing Language and Culture University</figcaption></figure><p><em>[30/08]</em> Our last evening in Beijing.. It has been lot’s of fun, but tomorrow, Shenzhen is awaiting us. First of all, let’s reflect upon these last few days. Three days ago, we had the chance to practice some material arts. This was amazingly fun! The teacher showed us a bit taichi and kung-fu and we tried (and mostly failed) our best to perform them well. After a good three hours of practicing, the lesson was over and we went to the K-tv. This is an unique Asian Karaoke. Once entered, we went into a room, that was close off just for us. We sang our hearts out for about two hours (and had no voice left afterwards). To me, this was one of the most fun experiences so far of this trip! After we had eaten, we went back to our hotel and prepared for our exam. As the exam consisted of two parts — oral and written exam — we decided to prepare for the oral part. We took it very seriously and made a little theatrical piece involving Chinese dialogues, kung-fu, some dancing and some singing. Afterwards, we studied some more and went sleeping.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=http%3A%2F%2Fscontent-a.cdninstagram.com%2Fhphotos-xaf1%2Ft50.2886-16%2F10663864_674262955991519_34215925_n.mp4&amp;src_secure=1&amp;url=http%3A%2F%2Finstagram.com%2Fp%2FsVOOttR_Se&amp;image=http%3A%2F%2Fphotos-f.ak.instagram.com%2Fhphotos-ak-xfa1%2F10643909_1563849813842821_1018020737_n.jpg&amp;key=d04bfffea46d4aeda930ec88cc64b87c&amp;type=video%2Fmp4&amp;schema=instagram" width="616" height="716" frameborder="0" scrolling="no"><a href="https://medium.com/media/eb1d7de5ccacc914e3761280066d545f/href">https://medium.com/media/eb1d7de5ccacc914e3761280066d545f/href</a></iframe><p>The next morning, we overslept a little bit because we were really tired (but we still made it in time for our lessons). We finished our Chinese textbook completely that morning. In the afternoon, we had the chance to do a cultural exchange with the local students (that knew English). The first part consisted of introducing ourselves (some in Chinese, some in English, some both!). For the second part, we had to show of some of our talents. Anna did some kung-fu, some of the boys sang Broeder Jacob (in English, Chinese, French and Dutch). They even thought the local students how to sing the Dutch version. I told all the students during the introduction that I knew how to play the drums and so they were expecting me to perform. Not having prepared for this, I just stood up, went to the front of the classroom and looked out for things that would make a reasonable sound! For the rest of my performance, you can see it here below ;)</p><p><a href="https://www.facebook.com/video.php?v=1537463719805960&amp;fref=nf">魏铭霈 - hahaha the guy was gonna smash the public... | Facebook</a></p><p>Later in the evening, we all were studying hard, so we would score great on our exam the day after! As it were <a href="http://instagram.com/p/sRVsFAR_YJ">more then 200 words</a>, we were up quite late studying! After another short night, we were facing our exams with lots of courage! Luckily for us, the studying payed of really well and we all really aced our exams! The local teachers said multiple times that, of all the 12 groups that were here with the seeds program (from different countries), that our performance or theater piece was the best one they had ever seen!</p><blockquote><strong>非常好！</strong> — (Very good)</blockquote><p>The teacher almost cried when we had to say goodbye to her… After we got supper at the VIP room in the university’s restaurant, we went to the cafeteria to spend our last money on our cards. We bought some <a href="http://instagram.com/p/sR-ZKfR_dF">drinks</a> and went back to the hotel. After a few hours of laughing and talking, we decided to go out to a nearby club, called the Wu-club. It was an amazing night! The music was very European, like in Belgian clubs, but the prizes for the drinks were way cheaper!</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/640/1*_5esGnbqSiE-CeVfkvKp7w.jpeg" /><figcaption>The most steep part of the whole Great Wall was the part we had planned to do.</figcaption></figure><p>After another really short night, we went to the Great Chinese Wall! It took us about two hours to get there. Once there, I was amazed and disappointed at the same time! It was amazing to see this great monument with my own eyes, but it was a really foggy day, so we couldn&#39;t really enjoy the full view of the wall. The slope was steep and the steps were really high, but it was something we just had to do! It would be weird, going to China and not having climbed the wall. Still, it was really hard for all of us, as we were all still a bit tired from the night before. After lunch, we had a long bus ride going to the <a href="http://instagram.com/p/sUvMHLx_Xa">Temple of Heaven</a>. In the old times, this temple was used by The Emperor to pray for a good harvest so he would only go there two times a year. Personally, I think it was an overwhelmingly big space that was used up, just to go there two times a year. As we counted, everything was back in nine or three. For example, the amount of steps leading to the next plateau were 9 steps and there were 3 plateaus to the temple or monuments. This is done a lot in China when having to do with The Emperor.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/640/1*q-gH_a5aTkFGej6jcxI6Wg.jpeg" /><figcaption>MK, LV, ROLEX, … You name it and Hongqiao market has it at an amazing price with good quality!</figcaption></figure><p>After we left the domain, we went to the Hongqiao Market (or as our guide would call it, the fake market). It’s an enormous complex with 5 floors, just filled with fake handbags, shoes, clothing, jewelry, watches, … I had an amazing time there! I loved to negotiate the price and to see how low it could drop. This was something really unique and I still talked about it to the others hours later, when we were sitting together in our room, watching some tv and realizing that this was our last night here. Tomorrow, we’re getting up early again and we’ll be flying off to Shenzhen in the afternoon!</p><h3>The real challenge starts now!</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/512/1*B8sP6CBYbijwJY--jsoFGQ.jpeg" /><figcaption>With it’s 18m above the ground and 6m below, this enormous wooden statue of Buddha is the biggest one in the whole wide world</figcaption></figure><p><em>[01/09]</em> Waking up in the morning, realizing we were leaving this beautiful place behind was weird and kinda sad, but the thought of going to another place filled with technology, opportunities and mystery overwhelmed these negative thoughts. Once we had our breakfast, we got in the bus to go to the Confucius temple. At first, I thought this was going to be just another temple, but it was so much more. In the old times of the dynasty’s, it was used as a school with very wise teachers. There was a <a href="http://instagram.com/p/sV-OzLx_d6">beautiful statue</a> of one the great Confucius himself in the garden. Afterwards, we went to the Lama temple, a Buddhist temple not that far from the temple we just went. There we saw several Buddha statues, each with a different meaning. At each one, we did our prayers and watched as the statues became more and more imposing. The greatest statue (as seen above) was really jaw dropping. It was just enormous! Leaving this temple behind, we headed to the airport. We had just one cultural sight left though. While we were in the bus, we passed the Olympic Stadium from the Beijing Olympic Games in 2008. <a href="http://instagram.com/p/sWOROiR_ae/">The Birds Nest</a>. It was imposing, even from such a distance! Arrived at the airport, we ate and said goodbye to our local Beijing guide Maria. About an hour later, we went on the plane and waited for it to take off. The problem though was that the weather in Shenzhen and Beijing was really bad, so it wasn&#39;t really safe to fly at that moment. So we waited for the ‘storm’ to pass… And waited… And waited even more… After about four hours, the plane finally left off, leaving us with our last gaze of Beijing.</p><blockquote>So long Beijing, we’ll miss you!</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/432/1*n78UvuvplSTmch3gbFOkhQ.jpeg" /></figure><p>About a small three hours later, we safely landed in Shenzhen. It was already dark there, so the <a href="http://instagram.com/p/sXSD83x_fl/">view from the plane</a> was pretty amazing! We were happy to get our suitcases rather quickly. Exiting the terminal, we met our Shenzhen local guide, called Laura. She led us to an Huawei bus, which we took to the Huawei campus to get a good nights rest.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*3ccQE_EJgtiXlmTM7nOsLA.jpeg" /><figcaption>A path that leads around the inner garden in the Huawei’s headquarters in Shenzhen</figcaption></figure><p>The day after, our first official day at Huawei started. For the next few weeks, we would learn all about their technology’s, about their solutions and about their way of thinking. To start the hot and tropical day off, we had a course to introduce us to the company and it’s (core) values. After this presentation was over, the teacher needed two volunteers to use in a stop motion commercial. Sanne and me were so excited, that they’d let us do it. Hopefully, the end result will be as great and beautiful as we are expecting. In the afternoon, we visited several area’s in Huawei. The most appealing area for me was the future area. It was a space where there were all sorts of presentations, predicting how the future for Huawei and the world would look like. For example, they are striving to get the internet of things connected to the physical world, while communicating through perhaps already 5G networks, with a setting date of only 2020! This was just one of the many things that amazed me today. I got a chance to see one of their data centers, their mobile cloud containers, their enterprise solutions, … For me, it was the extra excitement that made me happy and even more willing to learn about the technologies Huawei provides.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*CleSOtVs4njdGOnYSRALgw.jpeg" /><figcaption>Inside the Huawei company, showing and demonstrating their telecom solutions</figcaption></figure><p>After we had supper, we went to the swimming pool. When i talked before about the Huawei campus, I really meant the Bao Cai Garden. It’s an enormous domain, covering more then 100 000 m², with lot’s of nature. The swimming pool was scaled to these proportions. It has lot’s of place to play, relax and to swim (50m length). We ended our first day, watching one of the movie classics of all times, Pulp Fiction. For me, it was my first time seeing the movie, but I totally loved the dialogues and scenes. As it’s already late now, tomorrow will come quickly and I’ll be embracing it with open arms. Let the technology courses come!</p><h3>Cloud — Pipe — Device</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*JKZcoNcJtNwOmifT8_bx3A.jpeg" /><figcaption>Inside of the beautiful K-TV building</figcaption></figure><p><em>[03/09]</em> Wow! We almost can’t keep up with the speed or the courses we’re following! Yesterday, we had an introduction into Cloud, Pipe and Device. This is the basic architecture for every cloud solution. As the course went on, we got in depth more and more. For me, these lessons are just wonderful, as I’m making my way up to my masters degree (Cloud Computing). It’s just hard concentrating every second of the lesson, because if you don’t, you’ll end up not being able to know how we got to an exact point. After all our courses were done, we went back to the hotel and put on our swimming suits! The pool here is just amazing. We played our own special form of water basketball. This meant that we had to score the basketball in the goal. That was the only rule that applied and so, sometimes, we would go a little bit rough on each other. Not that we minded though! After our skin was a bit wrinkly from the water, we got out, took a shower, ate and went straight to the K-tv. We sang our lungs out and danced till our legs were immensely exhausted. Returning to the hotel, we all had a good nights rest. The next day, we had another day of courses. This time, it was about the 2G/3G/4G connections and transitions. It was really a lot to take in at once. In the evening, we ate at a Huawei restaurant, studied a bit and went to sleep…</p><h3>Three nights to remember</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/960/1*tjSkNH0UJILHKiOwRM-36A.jpeg" /><figcaption>A selfie taken at the crazy party night at the Face Club with McKidor</figcaption></figure><p><em>[06/09] </em>Finally, after a long time of hard courses, we are getting some time off!! The last three days were just amazingly wonderful! On thursday, we had some lab practice in the Huawei building. It involved 2G/3G/4G communication technology with firsthand Huawei devices as the BSC6900. As we quickly finished our exercises, Anna, Olivier, Jolien and me went for a coffee in the nearby building. While we were walking towards the shop, we heard some clouds colliding. A storm was coming. Unknowing of what would happen, we sat on the terrace, enjoying our coffee. After a while, an immense wind was present, taking water bottles, garbage bins and signs with it. We were still enjoying our coffee and ignored the weather. Suddenly, we saw lightning strike and almost immediately, we heard the clashing sound of clouds. It was deafening. Quickly we went inside the shop, where we witnessed a lot more lightning and heavy rain. As we had to leave with the bus soon, we asked for some garbage bags to but our shoes in and to pull over ourselves. That way, we made our way back to the hotel without getting wet too much.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*BFR4fMlLRssoEfty9N_AGA.jpeg" /><figcaption>A brief picture of the Shenzhen skyline, taken at a Chinese hotpot restaurant</figcaption></figure><p>After we had refreshed, we got the opportunity to go to the city center (for the first time) to attend a traditional Chinese opera. At first, we read about the story on wikipedia, so that we could follow along better. It was called the L<a href="http://en.wikipedia.org/wiki/Legend_of_the_White_Snake">eifeng Pagoda</a> or as translated, The legend of the White snake. It tells the story about a white snake becoming immortal and having the power to change herself into a human form. As for the opera itself, it was mindblowing! Of course, there were some moments of long dialogue that we could not follow, but still, The dancing, costumes and sword fighting amazed us so much. A real experience to never forget! After a nights rest, we had our last course at the Huawei training center. It involved how the Internet works and what the future of networking will bring. It was immensely interesting as I now finally understand some basic fundamentals regarding the IP-network. As a fun fact, I would say that our teacher for that day, named Mao Ping Gui, was family to the Great Mao Zedong. In fact, the father of Zedong was the brother of the great grandfather of Ping Gui! In the evening, we had a bus arranged to bring us to Dongmen Market. This was an immense shopping district, located in the center of Shenzhen. Once we were closing in on the market, the real Shenzhen became clear. A huge amount of lighted skyscrapers filled the Shenzhen skyline and we were amazed again by the greatness that China has to offer. Once we got out, we decided to split up and meet up again later at the Mc’Donalds. Oliver and me looked around a bit further up ahead and we came across a weird little store with lot’s and lot’s of anime and Kpop mugs, posters, statues, … Everything you could think off! I had an amazing time just looking through all the stuff they had lying there. When we met up again with the others, we recommended to go to the LunaPark we had passed on our way back. This way, my dream of dancing on a DDR station became reality. I danced, played some drums and percussion games till I literally couldn&#39;t take it anymore. My shirt was soaking wet from giving it my best.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*gjfPLmcqvhVOAhPe1FM5CQ.jpeg" /><figcaption>The LunaPark, with at the left, a DDR machine and on the outer right, a Roland Drum</figcaption></figure><p>Afterwards, we walked around and asked some foreigners where the best place to club would be. They firmly informed us that we should go to the Face-Club, which was reachable in a good fifteen minutes. Once there, we got to the fourth floor, where all people attending the party where just sitting at tables, drinking, playing games and talking. This is the way to party in China (with lot’s of champagne), but for us, it isn&#39;t. We wanted to take the party up to the next level, so we decided to go up front and party along with the<a href="http://instagram.com/p/slcTW7x_aF/"> dj’s.</a> They totally loved it! The McKidor gave us some wine, shook our hands and interacted with us once in a while. His mate, the other DJ, was a Dutch DJ (which we unfortunately did not know at the time). We had such a great time, but six of us, including me returned to the hotel a bit earlier then the rest. For us and for this night, we had had enough. Personally for me, it was one of the best nights of our whole China trip.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=http%3A%2F%2Fvideos-g-6.ak.instagram.com%2Fhphotos-ak-xfa1%2F10688678_549996981768926_194321998_n.mp4&amp;src_secure=1&amp;url=http%3A%2F%2Finstagram.com%2Fp%2FsnM2peR_R4&amp;image=http%3A%2F%2Fscontent-b.cdninstagram.com%2Fhphotos-xaf1%2Ft51.2885-15%2F10616286_638651129584551_1170167376_n.jpg&amp;key=d04bfffea46d4aeda930ec88cc64b87c&amp;type=video%2Fmp4&amp;schema=instagram" width="616" height="716" frameborder="0" scrolling="no"><a href="https://medium.com/media/724292de585be69efe9a8bde7854f193/href">https://medium.com/media/724292de585be69efe9a8bde7854f193/href</a></iframe><p>Waking up the next morning with the thought that we only had to get on the bus around one o’clock was wonderful. I turned around and slept some more. Around nine, Anna posted a message on our groupchat to ask who’d accompany her and Tom for breakfast and as I was wide awake, I got ready and went downstairs. Afterwards, we put on our swimming suits and took a little dive. The water felt enormously good and refreshing. We sunbathed a bit till noon and afterwards, we went shopping for BBQ-food. After our drive to the beach — which was a two hour drive — we had some fun in the ocean. It was a small area where we could swim, but it was big enough for all of us. As the night was making his entrance, the light of the sun was lowering like it was being dimmed by a switch. After a while, we got off the beach and started our great BBQ. The meat was nicely cooked and it tasted amazing. When the BBQ was over, we went for one last swim underneath the moonlight sky. It was really beautiful… If you ever get the chance, you should totally watch a Chinese opera, go clubbing in China and swim beneath the moonlight sky! I can recommend it!!</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*PchdrdAYlR9V5puxawpQNw.jpeg" /><figcaption>Francis is having some selfie fun at the beach</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*zZmfEy_PpuRf9UwqHr2_vg.jpeg" /><figcaption>BBQ, warm weather, beach, moonlight and amazing friends… This is the life</figcaption></figure><h3>Mid-autumn Festival</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*kNvE2Y7vR_ISbtiSbgXA-A.jpeg" /><figcaption>A view from a sunny day in the city center of Shenzhen, with in the middle, the 2nd highest building in the whole city</figcaption></figure><p><em>[8/09]</em> The end is nearing… We are feeling it and it comes to mind more and more, as we’re closing in on our last few days. The fact is though, we’re not done yet and we’ll surely make the best of it like the past weeks! Yesterday morning, we drove back to the city center to go shopping and sightseeing! In the morning, we went to the great fake market of Shenzhen. It was totally different as from the Beijing fake market and personally, I liked the one in Beijing way more. When we arrived, we saw a huge shopping complex with 5 floors that stretched out really far (picture below). The problem was though, that it was only 10:30 and most of the shops only opened at 11:00. As we walked through the building, just watching the shopping vitrines, we were followed by a few shop owners, trying to lure us into their shops. It was totally obnoxious. We just peacefully wanted to walk around a bit and enter e few shops. When we finally got rid of them, we went inside some of the shops. There, the great difference between Shenzhen market and Beijing market was clear. In Beijing, the shops were all little and they showed of their products. You could just pass, bargain and buy! The quality in Beijing was fine too if you looked good enough and the price was really cheap. In Shenzhen on the contrary, it were two extremes against each other. In some shops, you could see that the bags and wallets were fake from miles away. Here the price would drop really low, but who would buy this? In some other shops though, you would just see a few good quality products and you would get a catalogue to look in. You just had to select the wallet, handbag, … you wanted and they would get it for you. The quality and eye for detail were amazing for those products, but the price could not drop nearly as low as in Beijing.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*T2Lr1eJdMBv_RTm0CuFInw.jpeg" /><figcaption>The Shenzhen fake market, with five floors of little shops</figcaption></figure><p>After a little two hours inside, we went to eat and drove to traditional Chinese village (so they told us). When we arrived, we thought that the expression of ‘traditional Chinese village’ must have been lost in translation somewhere, because we were in some kind of ‘amusement’ park. It was full off little versions of the great monuments, buildings and temples that China is so proud of. The Great Wall, The Forbidden City and the Terracotta Army were just a few of the monuments we saw there, in their miniature form. You could compare it to Belgium’s “Mini Europa”. We also got to see a Mongolian horse riding show, which illustrated the battle between the Ming and the Qing Dynasty and a dancing show that lasted about an hour. Back at the hotel in the evening, we opened our books and learned a bit, as we had our test the following day!</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/750/1*Y-4bL2aRciBGRTYcspRdSw.jpeg" /><figcaption>I got a haircut in China for almost no money. It also included a head and neck massage!</figcaption></figure><p>Waking up after a good nights rest would have been amazing, but as the amount of sleeplessness was getting higher, it also got a bit under our skin. Still, after a good shower in the morning, I’m widely awake! When we arrived at the training center, we first had a last time of lab practice. We had to configure a DSLAM to let us get connected to the voice, IP and IPTV service. Afterwards, we had our little test, which was totally different from what we had expected. The questions were far too easy and it was an open book test. For example: “Which of these are not a social network”. Way below what we learned I think. In the afternoon, we practiced our song we had to perform tomorrow at the closing ceremony. The song is a traditional Chinese love song about the moon as a metaphor. It fits perfectly as today is the Moon Festival, also called the Mid-Autumn festival. After practice, some of us went to the city center and Tom, Anna, Jolien and me went to the hair saloon. It was a real experience for us. At first, they washed our hair for about half an hour, including a head and neck massage. Afterwards, I showed them a picture of George Clooney and his hairstyle and I wondered what they could do with my hair to make it like this. Fortunately, it worked out well! In the picture you are seeing above, my hair is a bit big and voluminous, but when I’m doing it now, it comes out great! In the evening we went swimming and walked around in the neighborhood. As we were still in our swimsuit, with tshirt and shorts (and towels), people were looking weird at us. Tom did not have the fortune of having his shorts with him, so he just walked around in tshirt and Speedo. All of the people we passed stared at him and laughed! For us, it was really fun as we were making jokes all the time. Back at the hotel in the evening, we met up with the others to ask what they had done. After a few great stories from both sides, we went for a good nights rest, because in the morning, we would have to face our Closing Ceremony… Wish us luck!</p><h3>The end has come…</h3><p><em>[13/09] </em>It has come.. The end of our amazing trip.. It’s just unbelievable that we are heading home again, after all we have experienced! It’s so sad, but we knew from the start is was going to end someday… So with this last post, I’m giving you a final view into the passed four days of our China trip.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*fv4QyjPsbQ6_M2wBorKwAQ.jpeg" /><figcaption>All of us in front of the Closing Ceremony board, full with pictures from during our trip</figcaption></figure><p>Tuesday, the big day! We had our first good nights rest in a long time! Except for Anna and Tom though. They had to get up early to prepare for their speech and presentation. After all, this was the day of the closing ceremony! I took a nice long shower, did my hair and put on my tie and suit. Classy and stunning looking, we all went to the reception room. There, we patiently awaited the Belgian Ambassador and his delegation. When everyone arrived and took seats, the speeches began. They were very inspiring, even the one from Anna and Tom. Afterwards, we got our certificates one by one and we were all really proud. Tom and Anna were extra proud though, because they had been selected as the outstanding two students. That way, they both got a Huawei Ascend P6. This was a huge step forward for Tom, as he still was using an old — but indestructible — Nokia. We also had to do a little performance for all the people attending so we sung Yue Liang Dai Biao Wo De Xin, whilst each holding his own rose. At the end, we all gave our rose to the attendees. As the closing ceremony came to an end, we were send to the exhibition hall, where we had to do our first of many interviews. This time, it was for KanaalZ and Trends Magazine. They asked us about the trip, what we had learned, … When that was finally over, we took a bus (arranged by our lovely guide Laura) to the city center! First, we went to a restaurant in a shopping mall. Like the real Chinese we are, we ordered many dishes and just tried them all out. The most exquisite dish we had ordered was turtle and I have to admit, it wasn&#39;t bad at all! All stuffed, we went inside the shopping mall, bought some drinks and sat outside underneath a little tree.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*xwDDjt7A_IEU0lxSL0mP5A.jpeg" /><figcaption>Underneath the tree, from the left to the right: Me, Tom, Francis, Klaas and Karel</figcaption></figure><p>There, we played a game called “I have never”. For those who are not familiar with the game, the concept is to say something you have never done before. That way, all the people who have done it before have to drink and have the chance to share their stories. It was really fun to hear all those crazy tales from the others! At around midnight, we decided to go to a club and so, we ended up at Lily Marleen. It was a crowded club, full of Chinese people, willingly to meet new friends. Some even bought us drinks (and even chicken). It really was an amazing night and we all had lots of fun! The next day, we visited one of the Huawei factories, called Songsan Lake Factory. There, they are manufacturing a part of the smartphones. I had always been wondering how this process worked and how they actually did the check ups. The amount of automation was big! I even saw an automated SMD line! After we got lunch, we got back to the Huawei campus. We had to do another interview with KanaalZ / Trends, but this time, we had to be at work in the test lab. Afterwards, we just spend some time in the hotel. I used my time to research the Huawei Honor 6. I first laid my eyes upon this ‘monster’ when we arrived at the Huawei HQ last week. When I heard the ridiculously low price tag, I was sold. I wanted to know what version to buy, as it had to work in Belgium too. In the evening, we just gathered in one of the rooms and talked, listened to some music and watched a part of “The Wolf of Wallstreet”.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*FVR-9Jdv_m6wlBvCABqcGg.jpeg" /><figcaption>A commercial for the Honor 6 and the Ascend P7</figcaption></figure><p>After another good nights rest, we put on our suit again for the last interview. This time, it was with Huawei People Magazine and we just had to answer some basic questions about our trip. He wanted us to share our stories. When I had finished my part, I went back to the hotel to out on some more comfortable clothes and went straight for the official Huawei Shop. Honor 6, welcome to the family!! In the afternoon we took the bus to the famous Shenzhen Electronic Market. It was the combination of official and unofficial shops. At some places, you could even buy spare parts for some smartphones. I personally looked out for some cheap smartphone accessories, as cases, screen protectors, adapters, … Around six, Laura came along and we drove to the KK100, the highest building in Shenzhen (11th highest of the world). There, we went all the way up to the 96th floor. The view over the skyline was just breathtaking. Combined with a Chinese girl that was playing piano in the background, the moment was perfect. After a few minutes, darkness came to the city and everywhere, lights popped up from the buildings and streets. So unbelievably beautiful. We also got to eat the enormous and delicious buffet on the 95th floor. Combined with the spectacular view (and that this was the 2nd highest restaurant of the world), the price tag was huge, but it seemed Huawei wanted to invest this capital in us. Once back in the hotel, Laura had prepared a little gift for us. The first one was a really slick and slim external battery. The second present was emotionally worth so much, as she gave us a photo album with lots of different pictures taken during the whole trip. We were all touched. Back in our hotel room, we all made sure that our luggage was as good as finished. A little bit after midnight, we went downstairs to have a quick drink on our past few weeks, but also on Pieter his 20th birthday! Not long afterwards, we all went to bed to get our last night of sleep in the great land of China…</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*h4Z6xkG3bIVfaCKG-Xcr2A.jpeg" /><figcaption>The Shenzhen skyline, as viewed from the KK100</figcaption></figure><p>Friday… It’s over… It’s done… It’s our very last day of an unbelievable experience. For starters, we all got some breakfast and checked out. Ready to go, we said goodbye to our lovely guide Laura. It was with lot’s of sadness in our hearts that we had to leave her. She had been like a little Chinese mother to us these past two weeks. We saw that it was at least as hard for her too, as we saw some tears rolling from her cheeks when we left for Hong Kong. The trip only took one hour and half. It’s just unthinkable how two city’s with populations around the 10 million can only be so far apart. Arriving at Hong Kong, we stepped into another bus and drove straight to the top of the immense city. The Victoria peak! It was the highest point of the whole city, where we could see another amazing skyline. We also at a nearby restaurant and celebrated Pieter his birthday with cake and presents!</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*q2AKx8pCrF9WXMK7-86jgQ.jpeg" /><figcaption>A selfie taken at the Victoria Peak with the new Honor 6. In the background, you see the amazing Hong Kong skyline and an idea of the could-have-been-better weather</figcaption></figure><p>After some local shopping, we got back into the bus and drove back into the city. There, we got on a little ferry who took us across the river. As the weather wasn&#39;t that great, we decided to do some more indoor activities. That way, we ended up in a big shopping mall, full of shops selling highly expensive brands. Some examples were Gucci, Omega, Rolex, Michael Kors, Uggs, … After an hour or so, the rain stopped and we walked down the Star Avenue. There were a lot of stars of Chinese actors. Sadly, I only recognized Jackie Chan and Bruce Lee their stars. Once we arrived at the end of the lane, we got back on the bus and drove into the city center. We stopped at some little streets, lighted up by the huge amount of neon boards from the local shops. We walked through a little market and ended up at our place to have dinner, somewhere in tent in the corner of a street. After everyone was finished, we got back on the bus and drove to the airport. As some of us were still a bit hungry and had a lot of time left, we ate some McDonalds. It tasted the same as in Belgium, but at the time, it was really good. I hadn&#39;t eaten something like this in a long time. A bit later, it was time to go to our gate and witness the enormous plane we were going to board.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*9dmv0XFXw8H4KUaTC-eZcA.jpeg" /><figcaption>The airbus A380. With it’s two floors, it is the biggest passenger plane in the world.</figcaption></figure><p>The Airbus A380. If you look at the picture above and compare the plane with the truck on the left, you can estimate how big the plane is. It was really comfortable and the 8-hour flight to Dubai passed quickly. There, we drank some coffee and after a good three hours in the terminal, we boarded another plane going to Belgium. This time, it was just a Boeing 777. We had the immense luck that there weren&#39;t many passengers on this plane. This way, we could take four seats for our own and just lay down, get some sleep and dream of the unbelievable journey of the past three weeks… It was amazing…</p><blockquote><strong>FAREWELL CHINA, UNTIL WE MEET AGAIN</strong></blockquote><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=df0d185101f" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>