<?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 Architect Yeung on Medium]]></title>
        <description><![CDATA[Stories by Architect Yeung on Medium]]></description>
        <link>https://medium.com/@architectyeung?source=rss-3f56e5f686f5------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*SHwKIzU4QKwnxJi2kqZVlQ.jpeg</url>
            <title>Stories by Architect Yeung on Medium</title>
            <link>https://medium.com/@architectyeung?source=rss-3f56e5f686f5------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Fri, 20 Nov 2026 23:20:09 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@architectyeung/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[True Love 真愛]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://architectyeung.medium.com/%E7%9C%9F%E6%84%9B-6f9c5881bfdd?source=rss-3f56e5f686f5------2"><img src="https://cdn-images-1.medium.com/max/2500/0*LBFiNucJXL78OW8Q.jpg" width="2500"></a></p><p class="medium-feed-snippet">People who enjoy playing tennis are probably familiar with terms like to serve, love all, and play.</p><p class="medium-feed-link"><a href="https://architectyeung.medium.com/%E7%9C%9F%E6%84%9B-6f9c5881bfdd?source=rss-3f56e5f686f5------2">Continue reading on Medium »</a></p></div>]]></description>
            <link>https://architectyeung.medium.com/%E7%9C%9F%E6%84%9B-6f9c5881bfdd?source=rss-3f56e5f686f5------2</link>
            <guid isPermaLink="false">https://medium.com/p/6f9c5881bfdd</guid>
            <category><![CDATA[couples]]></category>
            <category><![CDATA[true-love]]></category>
            <dc:creator><![CDATA[Architect Yeung]]></dc:creator>
            <pubDate>Sat, 21 Mar 2026 11:18:56 GMT</pubDate>
            <atom:updated>2026-03-21T11:21:33.032Z</atom:updated>
        </item>
        <item>
            <title><![CDATA[Kazuo Inamori’s Top 10 Quotes about Life Philosophy]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://architectyeung.medium.com/kazuo-inamoris-top-10-quotes-about-life-philosophy-84ceec939135?source=rss-3f56e5f686f5------2"><img src="https://cdn-images-1.medium.com/max/600/1*O5MYispc5QmeMcl6b9-KbQ.png" width="600"></a></p><p class="medium-feed-snippet">Kazuo Inamori Quotes 1:</p><p class="medium-feed-link"><a href="https://architectyeung.medium.com/kazuo-inamoris-top-10-quotes-about-life-philosophy-84ceec939135?source=rss-3f56e5f686f5------2">Continue reading on Medium »</a></p></div>]]></description>
            <link>https://architectyeung.medium.com/kazuo-inamoris-top-10-quotes-about-life-philosophy-84ceec939135?source=rss-3f56e5f686f5------2</link>
            <guid isPermaLink="false">https://medium.com/p/84ceec939135</guid>
            <category><![CDATA[top-10-quotes-about-life]]></category>
            <category><![CDATA[life-philosophy]]></category>
            <category><![CDATA[inamori-kazuo]]></category>
            <dc:creator><![CDATA[Architect Yeung]]></dc:creator>
            <pubDate>Fri, 31 Jan 2025 00:07:48 GMT</pubDate>
            <atom:updated>2025-01-31T00:07:48.987Z</atom:updated>
        </item>
        <item>
            <title><![CDATA[稻盛和夫: 10大經典語錄]]></title>
            <link>https://architectyeung.medium.com/%E7%A8%BB%E7%9B%9B%E5%92%8C%E5%A4%AB-10%E5%A4%A7%E7%B6%93%E5%85%B8%E8%AA%9E%E9%8C%84-80ce9061f444?source=rss-3f56e5f686f5------2</link>
            <guid isPermaLink="false">https://medium.com/p/80ce9061f444</guid>
            <category><![CDATA[生命哲學]]></category>
            <category><![CDATA[稻盛和夫]]></category>
            <category><![CDATA[life-philosophy]]></category>
            <dc:creator><![CDATA[Architect Yeung]]></dc:creator>
            <pubDate>Thu, 30 Jan 2025 04:38:56 GMT</pubDate>
            <atom:updated>2025-01-31T00:09:08.535Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*gDgwp3AIIT6sW1JHOd-J0g.jpeg" /></figure><p><strong>稻盛和夫 語錄1:</strong></p><p>錢沒了可以再賺；工作沒了可以再找；朋友沒了可以再交；愛情沒了可以再找。你生來就一無所有，何懼從頭再來。水流到絕境是瀑布，人走到絕境是重生。人生如同登山，有高峰就會有低谷，半山腰總是擁擠的，你得努力去山頂看看。人生這道題怎麼選都有遺憾，真正的強者是夜深人靜的時候，把心掏出來，自己縫縫補補，睡一覺起來又是信心百倍。人活著就該逢山開路遇水架橋，睡前放下一切，醒來便是重生。</p><p><strong>稻盛和夫 語錄2:</strong></p><p>當別人不需要你的時候，要學會收回熱情並禮貌退場。你可以躲在被窩裡哭，也可以喝酒喝到吐，但是你不可以拿起手機發不該發的信息，打不該打的電話，你要明白，你得不到的要適可而止，及時止損才是成熟的表現，一扇不願意開的門，一直敲是沒有禮貌的。</p><p><strong>稻盛和夫 語錄3:</strong></p><p>不要把自己活得像落難者，急著告訴所有人你的不幸，你的經歷在別人眼裡就是一個笑話。成長本就是孤立無援的過程，你要相信自己的人生必將燦爛。一切都由心造就，一個人的人生，就是他的思維產物，正所謂心不換物不至。</p><p><strong>稻盛和夫 語錄4:</strong></p><p>大起大落後，你會發現，你幫過的人不一定會幫你，但幫過你的人，他還會再幫你。落難而識人，泥濘而識馬，一個人是真心還是假意，不在嘴上而在心上。一份情是虛偽還是實際，不在平時而在風雨。</p><p><strong>稻盛和夫 語錄5:</strong></p><p>欺負你的人，因你的軟弱而來；欣賞你的人，因你的自信而來；不在乎你的人，因你的卑微而來；愛你的人，因你的自愛而來。你遇到的人都是被你的能量吸引而來，你是誰便會遇到誰。</p><p><strong>稻盛和夫 語錄6:</strong></p><p>一個人如果連自己的情緒都控制不了，即便給你整個世界，你也早晚會毀掉一切。當你成不了心態的主人，必會淪為情緒的奴隸，人偶爾都會遇到糟心的人和事，所有的情緒都是一種表達，當情緒一次次來臨之時，你只需要學會按下暫停鍵，切記，脾氣永遠不要大於本事，沒有收拾殘局的能力，就別放縱善變的情緒。</p><p><strong>稻盛和夫 語錄7:</strong></p><p>你不會的東西，覺得難的東西，一定不要躲，先搞明白後再精湛，你就比別人優秀。因為大部分人都捨不得花力氣去鑽研，自動被淘汰，所以你執著的努力就佔了大便宜。請相信相信的力量，選擇一件事就堅持去做，低頭趕路不問前程，你會被越來越多的人看見，也會被越來越多的人吸引。 如果有一天你的努力配得上你的夢想，那你的夢想一定不會辜負你的努力。</p><p><strong>稻盛和夫 語錄8:</strong></p><p>三年入行，五年懂行，十年稱王。其實每個行業都是艱難的，没有一個行業是容易的。寧願十年做一件事也不要一年做十件事。走得快的人未必一路遙遙領先，走得慢的人或許會收穫意外的驚喜，人生的關鍵不是速度也不是快慢，而是要找到屬於自己的節奏，在此之前，我們需要多一點耐心。熬得住就出眾，熬不住就出局，成功靠得不是豪言壯語，而是腳踏實地不斷的努力堅持。</p><p><strong>稻盛和夫 語錄9:</strong></p><p>人生的高度不在於你看過了多少事，而是你看清了多少事。心靈的寬度不是你認識了多少人，而是你包容了多少人。做人如山，望萬物而容萬物；做人似水，能進退而知進退；看清不如看輕，看透不如看淡。人生就是一場修行，修行貴在修心。世事多變，唯有重新出發向心而行，才能在漫長歲月裡，不糾結、不驚慌、不迷茫，一切始於心而終於心。</p><p><strong>稻盛和夫 語錄10:</strong></p><p>錢的背後是事，把事做到極致錢自然來；事的背後是人，把人做好了事自然成；人的背後是命，把生命的維度修好自有好命；生活的背後是道，有一個助人的心就是正道。只有內心強烈渴望去做一件事情，才可能取得成功。命由己造，福由心生。如果你明白了這些人生道理，就會有新的高度和認知。</p><blockquote>如果稻盛和夫這10大語錄讓你深有感觸，請分享給你身邊的人，讓我們一同成就更好的自己。</blockquote><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=80ce9061f444" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Qi-Gong, a good way for longevity]]></title>
            <link>https://architectyeung.medium.com/qi-gong-a-good-way-for-longevity-2b739ed48ef5?source=rss-3f56e5f686f5------2</link>
            <guid isPermaLink="false">https://medium.com/p/2b739ed48ef5</guid>
            <category><![CDATA[health]]></category>
            <category><![CDATA[wellbeing]]></category>
            <category><![CDATA[wellness]]></category>
            <category><![CDATA[longevity]]></category>
            <category><![CDATA[qigong]]></category>
            <dc:creator><![CDATA[Architect Yeung]]></dc:creator>
            <pubDate>Sun, 31 Dec 2023 22:05:09 GMT</pubDate>
            <atom:updated>2024-01-08T09:59:06.949Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*qXh-9PLvr8I6Qmztixq1rw.png" /></figure><p><strong>Qi-Gong</strong> is also known is ‘Chi Kung’, and is an exercise system which focuses on the cultivation of Qi. Its roots in ancient Chinese culture date back more than 4,000 years ago. Most often, practitioners use Qi Gong to remedy illness, promote health and longevity, and increase internal strength for martial purposes.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*3vPc669E-c7vRRllu-UOmQ.png" /><figcaption><strong>A Universally understood concept: “Use it or lose it”</strong></figcaption></figure><p>It is said that a healthy body has ample and free flowing “Qi” with the appropriate amount and quality of Qi (i.e. the internal energy flow) at each part of the body and each organ. It is thought that Qi is the energy flowing throughout the Universe.</p><blockquote>It is thought that when Qi is abundant and flowing smoothly the body is healthy and strong. <strong>Qi-gong is an exercise to help accumulate Qi (energy) and cause the energy to flow smoothly around the body. The following diagram explains the meaning of Qi and Gong and their co-relation.</strong></blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*jIDuP3H2eUFcxupnMOWzOA.png" /><figcaption><strong>QI and GONG Relationship</strong></figcaption></figure><p>Traditional Chinese texts have described Qi and the ways to manipulate it to improve health and strengthen the body.</p><p>Traditional Chinese Medicine TCM describes Qi-Gong as an internal practice of physical exercise and mental meditation techniques to manipulate Qi. But they usually practice external methods for added benefits also. The external methods are herbal and medicinal methods and body manipulation along the channels in a variety of ways, such as moxibustion, cupping, massage, acupressure, and acupuncture.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*dLsZqw3PEkc3hkJhKL9_Zg.png" /><figcaption><strong>Qi Gong is like Medicine</strong></figcaption></figure><h3>Qi-Gong Methods</h3><p>There are many different techniques of Qi-Gong. Some are as simple as controlling breathing and meditation or doing small repetitive motions with the body. Other types of Qi-Gong are suitable for martial arts exercise. You might see martial artists practicing in the public parks some kinds of Qi-Gong that are the intricate and fast martial arts routines. Some routines are appropriate for martial artists, and some are more appropriate just for improving health.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/648/0*wKvQThK365vEv0I9" /><figcaption>Examples</figcaption></figure><h3><strong>Practical Benefits</strong></h3><p>Qi-Gong and Tai Chi practitioners generally say that the exercises calm them and help them to feel better and “more energised”. They say that they feel more energetic after a period of exercise than before they started.</p><p>The rhythmic motions tone the muscles and might help people improve balance and stamina. People learn how to control their limbs and body better, and the crouching stances in many of the more physical styles help to develop stronger legs. These can keep our body in a healthy status internally and physically.</p><p>It may surprise you, but Qi-Gong routines are thought to not only improve health and cure diseases, but to extend the life span.</p><p>The movements in Qi-Gong can be very slow. Students raise their arms over their heads, rotate their wrists, lower their arms and then make circular motions with their hands. As soon as you get into it, the Qi begins to flow.</p><p>Emotions are also calmed, and people tend to live happier.</p><blockquote>To achieve health &amp; wellbeing, one cannot rely on medical treatment or medicine alone. The lifestyle, or way of life impacts on the level of health and wellbeing one can enjoy. A healthy, balanced lifestyle is the key to avoiding preventable illness. Longevity depends on three major things:</blockquote><blockquote>Nourishment</blockquote><blockquote>Exercise</blockquote><blockquote>Mentality</blockquote><p><strong>Simply put, Qi-Qong helps you achieve health &amp; wellbeing in areas of Exercise and Mentality. That results in increasing the longevity and quality of your life.</strong></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=2b739ed48ef5" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Top 10 Technology Trends in coming three years, from 2024 to 2026]]></title>
            <link>https://architectyeung.medium.com/top-10-technology-trends-in-coming-three-years-from-2024-to-2026-f9a34847a87d?source=rss-3f56e5f686f5------2</link>
            <guid isPermaLink="false">https://medium.com/p/f9a34847a87d</guid>
            <dc:creator><![CDATA[Architect Yeung]]></dc:creator>
            <pubDate>Fri, 29 Dec 2023 02:57:20 GMT</pubDate>
            <atom:updated>2024-01-08T09:58:53.838Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/650/1*n_Ttdb_mVQQC9kuW5755IQ.png" /><figcaption><strong>Hephaestus | Ἥφαιστος | 赫菲斯托斯</strong></figcaption></figure><p>What’s next for driving changes?</p><p>By this year end, it’s interesting to consider what the top technology trends in the coming three years (2024 to 2026). This helps us prepare for these potential opportunities. These trends are listed and explained as follows:</p><blockquote>Trend 1: Generative AI, done<br>Trend 2: Sustainable Technology, done<br>Trend 3: Cybersecurity, done<br>Trend 4: Quantum Computing, done<br>Trend 5: Data Science &amp; Machine Learning, done<br>Trend 6: Web 3.0 &amp; Metaverse <br>Trend 7 : Automation<br>Trend 8: Autonomous Vehicles, done<br>Trend 9: 5G and 6G Technologies, done<br>Trend 10: Bio-technology, done</blockquote><h3>Trend 1: Generative AI</h3><p>Generative Artificial Intelligence (Gen. AI) will definitely one of the most interesting technological trends in coming years. Generative AI refers to a class of artificial intelligence systems that are designed to generate contents, whether it be text, images, audio, or other forms of data. These systems use machine learning techniques, particularly deep learning, to understand and replicate patterns within the data they are trained on. One prominent type of generative AI is based on generative models.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/785/1*Z80xvvkwrTsZ38wJG1d7XA.png" /><figcaption><strong>Evolution of AI vs Time</strong></figcaption></figure><p>The recent developments in this trend include ChatGPT by OpenAI in 2022. This also triggered a dramatic increase in investor interest in Generative AI related companies or startups.. For example, Nvidia’s stock price rose dramatically in 2023. The overall Generative AI related companies out-performed the broad market substantially. Gartner projects A.I. will generate 10% of all data by 2025. It makes 2024 a year of major growth. It is expected to be an increasing trend.</p><p>Other technology giants like Microsoft, IBM, Amazon, IBM, Google, Alibaba, Baidu and Salesforce are joining the game and have made strong progress in this space. Startups like AlphaSense, Jasper, Hugging Face, AlphaSense, C3 AI, DeepMind, Databricks, Synthesis AI, Stability AI, Lightricks and Inflection are potential ones in this area.</p><p>Generative AI technology would induce breakthroughs and changes to the economy. According to Goldman Sachs’s report, Generative AI could increase Global Gross Domestic Product by 7% in the coming 10 years (2024 to 2034) and bring significant impacts on society and people.</p><h3>Trend 2: Sustainable Technology</h3><p>This trend includes green technology and climate technology. It’s also being highlighted as one of Gartner’s top strategic technology trends. Sustainable technology is a framework of digital solutions that drive Environmental, Social and Governance ESG outcomes.</p><p>Sustainable technology, also known as green or clean technology, refers to the development and use of products, processes, and services that have a minimal impact on the environment, both in their production and consumption. The goal of sustainable technology is to meet the needs of the present without compromising the ability of future generations to meet their own needs. This involves considering environmental, social, and economic factors in the design, production, and use of technology.</p><p><a href="https://www.gartner.com/en/articles/are-you-thinking-too-small-about-sustainable-technology">Gartner</a> predicts that by 2027, 25% of Chief Information Officer CIO compensation will be linked to sustainable technology impacts. <a href="https://hbr.org/2017/05/saving-the-planet-from-ecological-disaster-is-a-12-trillion-opportunity">Harvard Business Review</a> reports that saving the earth from ecological disaster is a $12 trillion opportunity. This is true even if we meet just four out of 60 sections, including food and agriculture, cities, energy and materials, and health and well-being.</p><p>Thinking about the sustainable development goals in coming years, it’s expected to be at the forefront including electric vehicles, sustainable home and business construction, green and clean energy, agricultural technologies and carbon capture.</p><p>Google, Apple, Meta, and Amazon<em> have launched their initiatives to be more </em>sustainable enterprises. Companies like Nestle and Unilever are widely admired when it comes to sustainable measure-taking. Unilever increased its market cap threefold in the past 10 years by being a carbon-neutral-friendly company. Other companies like are other public corporations that have taken initiatives to be considered as ESG-friendly companies.</p><p>Top startups in this field include Amp Robotics, Aurora Solar, Oxford PV, Verdigris, Grow a Wish, Carbon Clean Solutions, Blue Planet, Heliogen, Grove Collaborative, Agricool, TIPA, Genecis, and Source.</p><h3><em>Trend 3: Cybersecurity</em></h3><p><a href="https://cybersecurityventures.com/hackerpocalypse-cybercrime-report-2016/"><em>Cybersecurity Ventures</em></a><em> expects the cost of cybercrime to reach $10.5 trillion by 2025, indicating the importance of cybersecurity for businesses. In the face of this fast-growing threat, cybersecurity solutions designed to strengthen defences and provide us with a fighting chance are on high demands for enterprises. Cyber-attacks are becoming more sophisticated, so competition to bring new solutions leveraging breakthrough technologies like A.I. to market is intensifying.</em></p><p>It’s expected that cyberattacks will likely become more complicated and intensive, employing A.I. to enhance defensive strategies. The proliferation of connected devices is expected to exacerbate security concerns, given their inherent vulnerabilities. Experts anticipate that generative A.I. will play a key role in this changing landscape.</p><p><em>We will likely see the emergence of stringent cybersecurity regulations across the globe. As governments and major entities intervene more than ever, companies will need to navigate a complex maze of rules in their efforts to bolster their defence mechanisms. I believe that 2024 will be a year in which technological advancements in cybersecurity will intertwine deeply with regulatory, human, and A.I.-driven perspectives.</em></p><p><em>Automated threat management, cloud security, zero-trust architecture, identity management, behavioural analytics, cyber governance, endpoint protection, cybersecurity-as-a-service, blockchain security, and cybersecurity mesh will remain top cybersecurity innovations in 2024.</em></p><p>Key companies in this field include CrowdStrike, Zscaler, Okta, SentinelOne, Palo Alto Networks, Fortinet, Splunk, Akamai, and McAfee are some of the top global public company leaders. Startups including Cyware, Lacework, Deep Instinct, Orca Security, GitGuardian, Snyk, Abnormal Security, EclecticIQ, Tanium, Nozomi Networks, and Claroty will continue to lead cybersecurity innovation in coming years.</p><h3><em>Trend 4: Quantum Computing</em></h3><p>Quantum Computing will dominate large-scaled computing in the coming decade. It will find applications in compute-heavy areas such as artificial intelligence, cloud computing, cryptography, big data, genome sequencing, meteorology, material science, optimisation of complex systems and financial modelling etc.</p><p>Quantum Computing has long been the stuff of science fiction, and in coming three years, it is going to be a reality. This emerging technology has the potential of evolving computing power, enabling us to solve more sophisticated questions at a previously unimaginable speed and scale. According to <a href="https://www.fortunebusinessinsights.com/quantum-computing-market-104855">Fortune Business Insights</a>, the quantum computing market is projected to grow from $928.8 million in 2023 to $6.5 billion by 2030, with a CAGR of 32.1 percent during the forecast period.</p><p>IBM unveiled its most powerful 433 qubit processor quantum computer, named “Osprey,” at the IBM Summit in 2022. IBM is planning to release a new 1,121-qubit Condor processor soon. It will be the first on the planet to exceed 1,000 qubits. An Australian startup called Q-CTRL is a highly potential one which integrated their technologies with IBM, aiming to drive the applications on Quantum Computing in the market.</p><p>IBM, Microsoft, D-Wave Systems, IonQ, Rigetti, Intel, Honeywell, Google, Alibaba Group, Nvidia, and Baidu are some of the leading public company players in the quantum computing sector. Top startups in this sector include Q-CTRL, Pasqal, QC Ware, Zapata Computing, 1QBit, ColdQuanta, Atom Computing, and PsiQuantum.</p><h3>Trend 5: Data Science &amp; Machine Learning</h3><p>Data science will continue to be a rapidly evolving field with several ongoing trends like AI and Quantum Computing. The use of Artificial Intelligence AI and Machine Learning ML techniques within data science has been a significant trend. This includes the development of advanced algorithms, predictive modelling and automated decision-making systems.</p><ul><li>Explainable AI (XAI): As AI and ML models become more complex, there’s a growing emphasis on making these models interpretable and explainable. Understanding how these models arrive at specific conclusions is crucial for their ethical use.</li><li>Data Ethics and Privacy: With increasing concerns about data privacy and ethical considerations, there is a growing focus on implementing robust data governance practices and ensuring ethical use of data.</li><li>Automated Machine Learning (AutoML): The development of tools and platforms that automate the machine learning pipeline, from data preprocessing to model deployment, to make data science more accessible to non-experts.</li><li>Natural Language Processing (NLP): The application of NLP in data science has seen substantial growth. This includes sentiment analysis, language translation, and the processing of unstructured text data.</li><li>Advanced Analytics and Business Intelligence (BI): There is an increased integration of advanced analytics and BI tools to provide actionable insights to businesses. This involves the use of dashboards, data visualisation, and interactive reporting.</li><li>DataOps: Inspired by DevOps, DataOps emphasises collaboration and communication between data scientists, data engineers, and other stakeholders to streamline the end-to-end data lifecycle.</li><li>Quantum Computing: Although in its early stages, the exploration of quantum computing’s potential in solving complex data science problems is gaining attention.</li><li>Responsible AI: The emphasis on responsible and ethical AI practices, including fairness, accountability, and transparency, is growing. Organisations are increasingly aware of the social implications of their AI applications.</li></ul><p>Snowflake is a cloud-based data warehousing platform that enables organizations to store and analyze large volumes of data. It has gained popularity for its scalability and ease of use. Palantir specializes in big data analytics and is known for its platforms that help organizations make sense of complex and large-scale datasets. It has been involved in various industries, including government, finance, and healthcare. DataRobot focuses on automated machine learning, providing a platform that allows organisations to build and deploy machine learning models quickly. It aims to democratise AI and make it accessible to non-experts. Alteryx provides a platform for data blending and advanced analytics. It aims to simplify the data preparation process and empower analysts to build predictive models without extensive programming. C3.ai focuses on enterprise AI applications, offering solutions for industries such as energy, manufacturing, and healthcare. Their platform is designed to accelerate digital transformation through the use of AI. RapidMiner provides an integrated data science platform for building and deploying machine learning models. It offers a visual workflow environment for data preparation, modelling, and evaluation.</p><h3>Trend 6: Web 3.0 and the Metaverse</h3><p>In coming three years, Web 3.0 will gain further traction, opening avenues for new technologies, specifically the metaverse and other virtual worlds built for gaming, training, social interactions and business. Web 3.0 will push for the rapid adoption of the Metaverse across businesses as users expect to have a more personalised and rewarding online experience.</p><p>The digital frontiers of the Metaverse, Virtual Reality (VR), and Augmented Reality (AR) are poised for transformative advances. Propelled by affordable, advanced headsets, the VR and AR sectors are ready to expand; meanwhile, the Metaverse, a collective virtual shared space, is anticipated to redefine digital interactions.</p><p>From education to entertainment, immersive experiences will revolutionise learning, human-machine interaction and gameplay. Meanwhile, AR will innovate sectors like healthcare with precision diagnostics and retail with virtual showrooms. Generative A.I. is likely to enable tailor-made content creation within these realms. The convergence of VR, AR, Metaverse and Generative A.I. is likely to blur the lines between the physical and virtual life, creating a future dictated only by our imagination and creativity.</p><h3><em>Trend 7: Automation</em></h3><p>Industrial automation will continue to grow and innovate, driven by the convergence of the internet of things (IoT), edge computing, A.I., machine learning, and 5G/6G. In coming three years, we expect more predictive maintenance, real-time monitoring, interconnected shop floors, automated inventory control, real-time data analysis for logistics optimization, and demand forecasting with A.I. algorithms. Artificial intelligence, robotics, optimized logistics, streamlined transportation, and workflow automation will reduce timelines as well as costs. Innovations in supply chain management technologies, such as paperless transportation documents, will accelerate the flow of goods and reduce costs.</p><p>ABB, Denso, Siemens, Emerson Process Management, Rockwell Automation, Honeywell Process Solutions, Schneider Electric, Microsoft Supply Chain Platform, Yaskawa Electric, AWS Supply Chain, and Omron Automation are leading supply chain automation players today and will be innovating in 2024.</p><p>We will also experience automation to transform the modern workplace in 2024. Workplace automation will reduce inefficiency and redundancy and transform remote and hybrid work. From routine administrative tasks to complex decision-making processes, automation is streamlining operations and boosting efficiency. <a href="https://www.gartner.com/en/newsroom/press-releases/2021-04-28-gartner-forecasts-worldwide-hyperautomation-enabling-software-market-to-reach-nearly-600-billion-by-2022">Gartner</a> reports that by 2024, organisations will lower operational costs by 30 percent using hyper-automation.</p><p>UiPath, Salesforce, ServiceNow, Cognizant, DocuSign, and SAP are some of the top companies that have contributed to workplace and business automation.</p><h3><em>Trend 8: Autonomous Vehicles</em></h3><p>Autonomous Vehicles, one of the most exciting emerging technologies (requiring super big investment capital), are set to transform transportation in 2024 further. By eliminating the need for human drivers, autonomous vehicles are expected to improve safety, reduce traffic congestion, and increase mobility for millions of people in long run.</p><p>We expect to see autonomous vehicles advance further in sensor technology, machine learning, and connectivity. This will enable the vehicles to navigate complex environments and interact with other vehicles and infrastructure in real time, paving the way for a future where driving is no longer the norm.</p><p>One of the most promising applications of autonomous vehicles is in the field of mobility services. By providing on-demand transportation that is safe, efficient, and cost-effective, autonomous vehicles could help improve accessibility and mobility for people who are currently underserved by traditional transportation systems. It’s an innovative business model which can also drive the adoption of Autonomous Vehicles.</p><p>Companies leading global autonomous driving technology include Tesla, Rivian, Zoox, May Mobility, Momenta, Pony.ai, General Motors, Nvidia, and Waymo (acquired by Google). According to Global Data’s recent prediction, the autonomous vehicle (AV) industry will not develop a fully self-driving car until 2035. However, we already see that GM-backed Cruise and Google-run Waymo have released fully automated robo-taxi fleets in San Francisco and in other U.S. states. I expect that 2024 will be a good year to watch the progress in this area.</p><h3><em>Trend 9: 5G and 6G Technologies</em></h3><p>5G and 6G technologies are expected to evolve and change the business landscape. A key factor is that the IEEE’s amended standard is expected in May 2024. It will provide device manufacturers with design specifications to govern interoperability and performance. It’s important that 6G is not yet a functioning technology. While some vendors are investing in the next-generation wireless standard, industry specifications for 6G-enabled products remain years away.</p><p><em>But 5G networks are enabling faster and more reliable connectivity. These are the latest generation of wireless technology, offering faster speeds, lower latency, and more reliable connectivity than in the past. With the ability to handle massive amounts of data at lightning-fast speeds, 5G networks still have the potential to revolutionise the way we are using technologies.</em></p><p>We can see that 5G networks will become even more prevalent in coming years, with advancements in areas such as edge computing, Internet of Things (IoT), and Virtual and Augmented Reality VR and AR. This enables 5G networks to become more intelligent and capable, providing faster and more responsive services to users along the way.</p><p>One of the most promising applications of 5G networks is in the field of Virtual and Augmented Reality VR and AR. With the high-speed connectivity provided by 5G networks, users can experience immersive and interactive virtual and augmented reality environments with ease.</p><h3><em>Trend 10: Bio-technology</em></h3><p>Bio-technology is still an emerging field that combines biology and technology to create new products and processes that improve human lives. From health care to agriculture, biotechnology has the potential to evolve industries and solve some of the world’s biggest challenges. In coming three years, we expect to see biotechnology become even more advanced, in areas such as gene editing, synthetic biology, and personalised medicine.</p><p>It’s expected to see several significant areas of biotechnology to grow: personalised medicine, gene editing and Crispr diagnostics, machine learning and A.I., stem cell technology, tissue engineering and bio-printing, big data and drug research etc.</p><p>Thriving startup Strand Therapeutics is developing the first platform for creating programmable, long-acting mRNA therapeutics capable of delivering potentially curative treatments. 23andMe offers DNA analysis with easy-to-understand results, while Color is a health care delivery platform that offers cancer prevention, screening programs, and health care services.</p><p>In addition to health care, biotechnology can transform industries such as agriculture and energy by developing new crops and fuels that are more efficient and sustainable. It’s clear that biotechnology is set to become an increasingly important technology trend, providing new solutions to some of the world’s biggest challenges.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=f9a34847a87d" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Arch. Series: How to design Twitter]]></title>
            <link>https://architectyeung.medium.com/twitter-architecture-a6f37335bc95?source=rss-3f56e5f686f5------2</link>
            <guid isPermaLink="false">https://medium.com/p/a6f37335bc95</guid>
            <category><![CDATA[education]]></category>
            <category><![CDATA[design]]></category>
            <category><![CDATA[technology]]></category>
            <dc:creator><![CDATA[Architect Yeung]]></dc:creator>
            <pubDate>Mon, 25 Dec 2023 07:35:41 GMT</pubDate>
            <atom:updated>2023-12-27T12:23:39.700Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*JjEilnvAPNK5mRt4xdtwdw.png" /></figure><p><em>We use this case to learn more about how to design a better architecture. The following content will cover key areas including: <br>- Requirements<br>- Estimation &amp; Constraints<br>- Data Model Design<br>- Database Engine Consideration<br>- API Design<br>- Architecture Design<br>- Detailed Technical Design</em></p><h3>Part I. Twitter</h3><p>Twitter is a social media service where users can read or post short messages (up to 280 characters) called tweets. It is available on the web and mobile platforms such as Android and iOS. Twitter is widely used by individuals, businesses, celebrities, and organisations for various purposes, including sharing news, engaging with audiences, marketing, networking, and expressing opinions. It has become a powerful platform for real-time information dissemination and public conversation.</p><h3>Part II. Requirements</h3><h4>Functional Requirements</h4><ul><li>Should be able to post new tweets (can be text, image, video, etc.).</li><li>Should be able to follow other users.</li><li>Should have a newsfeed feature consisting of tweets from the people the user is following.</li><li>Should be able to search tweets.</li></ul><h4>Non-Functional Requirements</h4><ul><li>High availability with minimal latency.</li><li>The system should be scalable and efficient.</li></ul><h4>Extended Requirements</h4><ul><li>Metrics and analytics.</li><li>Retweet functionality.</li><li>Favorite tweets.</li></ul><h3>Part III. Estimation &amp; Constraints</h3><h4>Traffic</h4><p>This will be a read-heavy system, let us assume we have 1 billion total users with 200 million daily active users (DAU), and on average each user tweets 5 times a day. This gives us 1 billion tweets per day.</p><p><em>200 millions × 5 tweets =1 billion per day</em></p><p>Tweets can also contain media such as images, or videos. We can assume that 10 percent of tweets are media files shared by the users, which gives us additional 100 million files we would need to store.</p><p><em>10% × 1 billion = 100 million per day</em></p><p>What would be Requests Per Second (RPS) for our system?</p><p>1 billion requests per day translate into 12k requests per second.</p><p><em>1 billion / (24 hours × 3,600 seconds) = around 12k requests per second</em></p><h4>Storage</h4><p>If we assume each message on average is 100 bytes, we will require about 100 GB of database storage every day.</p><p><em>1 billion × 100 bytes = around 100 GB</em></p><p>We also know that around 10 percent of our daily messages (100 million) are media files per our requirements. If we assume each file is 50 KB on average, we will require 5 TB of storage every day.</p><p><em>100 million × 50 KB =5 TB per day</em></p><p>And for 10 years, we will require about 19 PB of storage.</p><p><em>(5 TB + 0.1 TB) × 365 days × 10 years = around 19</em></p><h4>Bandwidth</h4><p>As our system is handling 5.1 TB of ingress every day, we will require a minimum bandwidth of around 60 MB per second.</p><p><em>5.1 TB / (24 hours × 3600 seconds )= around 60 MB per second</em></p><blockquote><strong>High-level Estimation Summary:</strong></blockquote><blockquote>Daily active users (DAU): 100 million</blockquote><blockquote>Requests per second (RPS): 12K per second</blockquote><blockquote>Storage (per day): around 5.1 TB</blockquote><blockquote>Storage (10 years): around 19 PB</blockquote><blockquote>Bandwidth: around 60 MB/s</blockquote><h3>Part IV. Data Model Design</h3><p>This is the general data model which reflects our requirements.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/918/1*ZE4Pkh0IsglqTT8572o6JA.png" /></figure><p>We have the following tables:</p><p><strong>users</strong></p><p>This table will contain a user’s information such as name, email, dob, and other details.</p><p><strong>tweets</strong></p><p>As the name suggests, this table will store tweets and their properties such as type (text, image, video, etc.), content, etc. We will also store the corresponding userID.</p><p><strong>favorites</strong></p><p>This table maps tweets with users for the favorite tweets functionality in our application.</p><p><strong>followers</strong></p><p>This table maps the followers and <a href="https://en.wiktionary.org/wiki/followee">followees</a> as users can follow each other (N:M relationship).</p><p><strong>feeds</strong></p><p>This table stores feed properties with the corresponding userID.</p><p><strong>feeds_tweets</strong></p><p>This table maps tweets and feed (N:M relationship).</p><h3>Part V. Database Engine Consideration</h3><p>While our data model seems quite relational, we don’t necessarily need to store everything in a single database, as this can limit our scalability and quickly become a bottleneck.</p><p>We will split the data between different services each having ownership over a particular table. Then we can use a relational database such as <a href="https://www.postgresql.org/">PostgreSQL</a> or a distributed NoSQL database such as <a href="https://cassandra.apache.org/_/index.html">Apache Cassandra</a> for our use case.</p><h3>Part VI. API design</h3><p>Let us do a basic API design for our services:</p><h4>Post a Tweet</h4><p>This API will allow the user to post a tweet on the platform.</p><pre>postTweet(userID: UUID, content: string, mediaURL?: string): boolean</pre><p>Parameters</p><p>User ID (UUID): ID of the user.</p><p>Content (string): Contents of the tweet.</p><p>Media URL (string): URL of the attached media <em>(optional)</em>.</p><p>Returns</p><p>Result (boolean): Represents whether the operation was successful or not.</p><h4>Follow or Unfollow a User</h4><p>This API will allow the user to follow or unfollow another user.</p><pre>follow(followerID: UUID, followeeID: UUID): boolean<br>unfollow(followerID: UUID, followeeID: UUID): boolean</pre><p>Parameters</p><p>Follower ID (UUID): ID of the current user.</p><p>Followee ID (UUID): ID of the user we want to follow or unfollow.</p><p>Media URL (string): URL of the attached media <em>(optional)</em>.</p><p>Returns</p><p>Result (boolean): Represents whether the operation was successful or not.</p><h4>Get Newsfeed</h4><p>This API will return all the tweets to be shown within a given newsfeed.</p><pre>getNewsfeed(userID: UUID): Tweet[]</pre><p>Parameters</p><p>User ID (UUID): ID of the user.</p><p>Returns</p><p>Tweets (Tweet[]): All the tweets to be shown within a given newsfeed.</p><h3>Part VI. Architecture Design</h3><h4>Microservice Architecture Pattern</h4><p>We will be using microservices architecture since it will make it easier to horizontally scale and decouple our services. Each service will have ownership of its own data model. Let’s try to divide our system into some core services.</p><p><strong>User Service</strong></p><p>This service handles user-related concerns such as authentication and user information.</p><p><strong>Newsfeed Service</strong></p><p>This service will handle the generation and publishing of user newsfeeds. It will be discussed in detail separately.</p><p><strong>Tweet Service</strong></p><p>The tweet service will handle tweet-related use cases such as posting a tweet, favorites, etc.</p><p><strong>Search Service</strong></p><p>The service is responsible for handling search-related functionality. It will be discussed in detail separately.</p><p><strong>Media service</strong></p><p>This service will handle the media (images, videos, files, etc.) uploads. It will be discussed in detail separately.</p><p><strong>Notification Service</strong></p><p>This service will simply send push notifications to the users.</p><p><strong>Analytics Service</strong></p><p>This service will be used for metrics and analytics use cases.</p><p>What about inter-service communication and service discovery?</p><p>Since our architecture is microservices-based, services will be communicating with each other as well. Generally, REST or HTTP performs well but we can further improve the performance using gRPC which is more lightweight and efficient.</p><p>Service discovery is another thing we will have to take into account. We can also use a service mesh that enables managed, observable, and secure communication between individual services.</p><h4>Newsfeed</h4><p>When it comes to the newsfeed, it seems easy enough to implement, but there are a lot of things that can make or break this feature. So, let’s divide our problem into two parts:</p><p><strong>Generation</strong></p><p>Let’s assume we want to generate the feed for user A, we will perform the following steps:</p><ol><li>Retrieve the IDs of all the users and entities (hashtags, topics, etc.) user A follows.</li><li>Fetch the relevant tweets for each of the retrieved IDs.</li><li>Use a ranking algorithm to rank the tweets based on parameters such as relevance, time, engagement, etc.</li><li>Return the ranked tweets data to the client in a paginated manner.</li></ol><p>Feed generation is an intensive process and can take quite a lot of time, especially for users following a lot of people. To improve the performance, the feed can be pre-generated and stored in the cache, then we can have a mechanism to periodically update the feed and apply our ranking algorithm to the new tweets.</p><p><strong>Publishing</strong></p><p>Publishing is the step where the feed data is pushed according to each specific user. This can be a quite heavy operation, as a user may have millions of friends or followers. To deal with this, we have three different approaches:</p><ul><li><strong>Pull Model (or Fan-out on load)</strong></li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*yjRgD_GkBGVjnEM7fVf2AQ.png" /></figure><p>When a user creates a tweet, and a follower reloads their newsfeed, the feed is created and stored in memory. The most recent feed is only loaded when the user requests it. This approach reduces the number of write operations on our database.</p><p>The downside of this approach is that the users will not be able to view recent feeds unless they “pull” the data from the server, which will increase the number of read operations on the server.</p><ul><li><strong>Push Model (or Fan-out on write)</strong></li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*6PGIHjonpcX-468NLI48Uw.png" /></figure><p>In this model, once a user creates a tweet, it is “pushed” to all the follower’s feeds immediately. This prevents the system from having to go through a user’s entire followers list to check for updates.</p><p>However, the downside of this approach is that it would increase the number of write operations on the database.</p><ul><li><strong>Hybrid Model</strong></li></ul><p>A third approach is a hybrid model between the pull and push model. It combines the beneficial features of the above two models and tries to provide a balanced approach between the two.</p><p>The hybrid model allows only users with a lesser number of followers to use the push model. For users with a higher number of followers such as celebrities, the pull model is used.</p><h4>Ranking Algorithm</h4><p>As we discussed, we will need a ranking algorithm to rank each tweet according to its relevance to each specific user.</p><p>For example, Facebook used to utilise an <a href="https://en.wikipedia.org/wiki/EdgeRank">EdgeRank</a> algorithm. Here, the rank of each feed item is described by:</p><blockquote><em>Rank = Affinity x Weight x Decay</em></blockquote><p>Where,</p><p>Affinity: is the &quot;closeness&quot; of the user to the creator of the edge. If a user frequently likes, comments, or messages the edge creator, then the value of affinity will be higher, resulting in a higher rank for the post.</p><p>Weight: is the value assigned according to each edge. A comment can have a higher weightage than likes, and thus a post with more comments is more likely to get a higher rank.</p><p>Decay: is the measure of the creation of the edge. The older the edge, the lesser will be the value of decay and eventually the rank.</p><p>Nowadays, algorithms are much more complex and ranking is done using machine learning models which can take thousands of factors into consideration.</p><h4>Retweets</h4><p>Retweets are one of our extended requirements. To implement this feature, we can simply create a new tweet with the user id of the user retweeting the original tweet and then modify the type enum and content property of the new tweet to link it with the original tweet.</p><p>For example, the type enum property can be of type tweet, similar to text, video, etc and content can be the id of the original tweet. Here the first row indicates the original tweet while the second row is how we can represent a retweet.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*qwiS-BAseNfzhnfyC2-Jxg.png" /></figure><p>This is a very basic implementation. To improve this we can create a separate table itself to store retweets.</p><h4>Search</h4><p>Sometimes traditional DBMS are not performant enough, we need something which allows us to store, search, and analyze huge volumes of data quickly and in near real-time and give results within milliseconds. <a href="https://www.elastic.co/">Elasticsearch</a> can help us with this use case.</p><p><a href="https://www.elastic.co/">Elasticsearch</a> is a distributed, free and open search and analytics engine for all types of data, including textual, numerical, geospatial, structured, and unstructured. It is built on top of <a href="https://lucene.apache.org/">Apache Lucene</a>.</p><p>How do we identify trending topics?</p><p>Trending functionality will be based on top of the search functionality. We can cache the most frequently searched queries, hashtags, and topics in the last N seconds and update them every M seconds using some sort of batch job mechanism. Our ranking algorithm can also be applied to the trending topics to give them more weight and personalize them for the user.</p><h4>Notifications</h4><p>Push notifications are an integral part of any social media platform. We can use a message queue or a message broker such as <a href="https://kafka.apache.org/">Apache Kafka</a> with the notification service to dispatch requests to <a href="https://firebase.google.com/docs/cloud-messaging">Firebase Cloud Messaging (FCM)</a> or <a href="https://developer.apple.com/documentation/usernotifications">Apple Push Notification Service (APNS)</a> which will handle the delivery of the push notifications to user devices.</p><h3>Part VII. Detailed Technical Design</h3><h4>Data Partitioning</h4><p>To scale out our databases we will need to partition our data. Horizontal partitioning (aka Sharding) can be a good first step. We can use partitions schemes such as:</p><ul><li>Hash-Based Partitioning</li><li>List-Based Partitioning</li><li>Range Based Partitioning</li><li>Composite Partitioning</li></ul><p>The above approaches can still cause uneven data and load distribution, we can solve this using Consistent hashing.</p><h4>Mutual Friends</h4><p>For mutual friends, we can build a social graph for every user. Each node in the graph will represent a user and a directional edge will represent followers and followees. After that, we can traverse the followers of a user to find and suggest a mutual friend. This would require a graph database such as <a href="https://neo4j.com/">Neo4j</a> and <a href="https://www.arangodb.com/">ArangoDB</a>.</p><p>This is a pretty simple algorithm, to improve our suggestion accuracy, we will need to incorporate a recommendation model which uses machine learning as part of our algorithm.</p><h4>Metrics and Analytics</h4><p>Recording analytics and metrics is one of our extended requirements. As we will be using <a href="https://kafka.apache.org/">Apache Kafka</a> to publish all sorts of events, we can process these events and run analytics on the data using <a href="https://spark.apache.org/">Apache Spark</a> which is an open-source unified analytics engine for large-scale data processing.</p><h4>Caching</h4><p>In a social media application, we have to be careful about using cache as our users expect the latest data. So, to prevent usage spikes from our resources we can cache the top 20% of the tweets.</p><p>To further improve efficiency we can add pagination to our system APIs. This decision will be helpful for users with limited network bandwidth as they won’t have to retrieve old messages unless requested.</p><p>Which cache eviction policy to use?</p><p>We can use solutions like <a href="https://redis.io/">Redis</a> or <a href="https://memcached.org/">Memcached</a> and cache 20% of the daily traffic but what kind of cache eviction policy would best fit our needs?</p><p><a href="https://en.wikipedia.org/wiki/Cache_replacement_policies#Least_recently_used_(LRU)">Least Recently Used (LRU)</a> can be a good policy for our system. In this policy, we discard the least recently used key first.</p><p>How to handle cache miss?</p><p>Whenever there is a cache miss, our servers can hit the database directly and update the cache with the new entries.</p><h4>Media Access and Storage</h4><p>As we know, most of our storage space will be used for storing media files such as images, videos, or other files. Our media service will be handling both access and storage of the user media files.</p><p>But where can we store files at scale? Well, object storage is what we’re looking for. Object stores break data files up into pieces called objects. It then stores those objects in a single repository, which can be spread out across multiple networked systems. We can also use distributed file storage such as HDFS or GlusterFS.</p><h4>Content Delivery Network (CDN)</h4><p>Content Delivery Network (CDN) increases content availability and redundancy while reducing bandwidth costs. Generally, static files such as images, and videos are served from CDN. We can use services like <a href="https://aws.amazon.com/cloudfront">Amazon CloudFront</a> or <a href="https://www.cloudflare.com/cdn">Cloudflare CDN</a> for this use case.</p><h4>Identify and Resolve Bottlenecks</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*6KFwi0beVRzIYgJ3.png" /></figure><p>Let us identify and resolve bottlenecks such as single points of failure in our design:</p><ul><li>“What if one of our services crashes?”</li><li>“How will we distribute our traffic between our components?”</li><li>“How can we reduce the load on our database?”</li><li>“How to improve the availability of our cache?”</li><li>“How can we make our notification system more robust?”</li><li>“How can we reduce media storage costs”?</li></ul><p>To make our system more resilient we can do the following:</p><ul><li>Running multiple instances of each of our services.</li><li>Introducing load balancers between clients, servers, databases, and cache servers.</li><li>Using multiple read replicas for our databases.</li><li>Multiple instances and replicas for our distributed cache.</li><li>Exactly once delivery and message ordering is challenging in a distributed system, we can use a dedicated message broker such as <a href="https://kafka.apache.org/">Apache Kafka</a> or <a href="https://nats.io/">NATS</a> to make our notification system more robust.</li><li>We can add media processing and compression capabilities to the media service to compress large files which will save a lot of storage space and reduce cost.</li></ul><h3>Part VIII. Reference</h3><p>TBD</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=a6f37335bc95" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[7 Key Aspects of Cyber Security for Organisations]]></title>
            <link>https://architectyeung.medium.com/7-key-aspects-of-cyber-security-for-organisations-aba94bb9b5f?source=rss-3f56e5f686f5------2</link>
            <guid isPermaLink="false">https://medium.com/p/aba94bb9b5f</guid>
            <category><![CDATA[technology]]></category>
            <category><![CDATA[cybercrime]]></category>
            <category><![CDATA[data-protection]]></category>
            <category><![CDATA[small-business]]></category>
            <category><![CDATA[cybersecurity]]></category>
            <dc:creator><![CDATA[Architect Yeung]]></dc:creator>
            <pubDate>Wed, 26 Jan 2022 14:06:05 GMT</pubDate>
            <atom:updated>2022-01-26T14:06:05.529Z</atom:updated>
            <content:encoded><![CDATA[<p>This article aims to highlight seven key aspects which can help organisations set up their practices / controls to cope with the increasingly complex cyber security threats.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*xa3V4TjtRrslGN_1cHo3Eg.jpeg" /></figure><h3>Aspect 1: Security Policy and Security Management</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/840/1*42wEbtAyZ9UETWW4fUnAmA.jpeg" /></figure><p>Security Policy is an important document in an organisation. It dictates security requirements and attitude of senior management with respect to cybersecurity risk management. Senior management should setup a mechanism to maintain and disseminate the requirements of<br>security policy to staff in a regularly basis.</p><blockquote>Checklist for Aspect 1:</blockquote><blockquote>✓ Staff should be given a chance to read through the security policy, understand security requirements of the organisation and acknowledge<br>to conform when they onboard.</blockquote><blockquote>✓ The policy should be put in somewhere the staff can refer easily.</blockquote><blockquote>✓ Policy should be updated and let the staff to re-acknowledge the policy regularly.</blockquote><h3>Aspect 2: Endpoint Security</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*aN5KWrPfCesYQhLQilVVFA.png" /></figure><p>Endpoint refers to personal computers or notebook computers used by staff to access business information during work. Email communication, web browsing and other business applications are all run on endpoints. Attackers would like to compromise the endpoint since it can be used as an entry point to access valuable information assets of the organisation.</p><blockquote>Checklist for Aspect 2:</blockquote><blockquote>✓ Endpoint computers should be protected by security software like anti-virus and anti-malware software.</blockquote><blockquote>✓ Signatures and security software should be kept up-to-dated to protect the endpoint from most recent threats.</blockquote><blockquote>✓ Security patches for endpoint computer operating system should also be kept up-to-date.</blockquote><blockquote>✓ IT staff should monitor the update status of the endpoints as well.</blockquote><blockquote>✓ User accounts on endpoint should be non-privileged (not Administrator)</blockquote><blockquote>✓ Proxy server used to filter malicious URLs during web browsing.</blockquote><h3>Aspect 3: Network Security</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/870/1*v8Jvdw_Vyw-A8PE27AJZGQ.jpeg" /></figure><p>Most organisations would make use of Internet to facilitate business information exchange. Internet connection inherits network security risks that external attackers may intrude the organisation network from outside. Firewall, Internet facing servers and other network devices should be configured properly to avoid intrusion.</p><blockquote>Checklist for Aspect 3:</blockquote><blockquote>✓ Firewall should be configured properly that minimise network ports of organisation network exposing to the Internet.</blockquote><blockquote>✓ Default rule on firewall should be “DENY”. Only “ALLOW” certain traffic based on business needs.</blockquote><blockquote>✓ Do not allow ANY from internal network to have access to Internet. Only allow approved IP addresses to have Internet access instead.</blockquote><blockquote>✓ Do not allow remote access (e.g. RDP) from Internet to internal servers.</blockquote><blockquote>✓ Firewall rules should be reviewed regularly.</blockquote><h3>Aspect 4: System Security</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/800/1*mnT7weTujRNTstaNSBtdtw.jpeg" /></figure><p>Organisations make use of information systems to process business<br>information. Some systems (e.g. web servers) are open to Internet to provide/collect information to/from the Internet. These systems are target of<br>attackers since the information the systems contained are valuable. System security guidelines and practices should be developed for mission critical systems.</p><blockquote>Checklist for Aspect 4:</blockquote><blockquote>✓ Password policy should be configured such that passwords of server should meet minimum length and complexity requirement.</blockquote><blockquote>✓ Servers should be configured securely (called hardened) with security policies enabled and unused services disabled.</blockquote><blockquote>✓ System patches should be updated timely to protect from recent threats.</blockquote><blockquote>✓ Internet facing servers should avoid storing sensitive information. Sensitive<br>information should be masked or encrypted when stored in servers.</blockquote><blockquote>✓ Input from Internet users (e.g. web server forms) should be filtered properly in application to avoid SQL Injection type of attack.</blockquote><blockquote>✓ For critical systems serving the public and performing critical missions, periodical penetration test should be performed by professional parties.</blockquote><h3>Aspect 5: Security Monitoring</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*uo6JfRktq_-Cwa4a_tWX7g.jpeg" /></figure><p>There is no way to ensure 100% security of endpoints, servers and network. Organisations should setup mechanism to monitor and detect if something suspicious is happening in information systems. The earlier a threat is identified, the earlier actions can be taken. The potential damage of the threat can then be minimised.</p><blockquote>Checklist for Aspect 5:</blockquote><blockquote>✓ Logging should be enabled in network devices (e.g. firewall) and servers.</blockquote><blockquote>✓ Logs should be centralised somewhere within the organisation for periodical review and monitoring.</blockquote><blockquote>✓ Review of the logs should be timely such that detected issues are taken care properly.</blockquote><blockquote>✓ Network traffic (e.g. Internet traffic) should be monitored to detect if any abrupt change in traffic pattern.</blockquote><h3>Aspect 6: Incident Handling</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*YoopLdtSyps-SjlNUcwcxQ.png" /></figure><p>System outages due to system issues or security incidents are not 100% avoidable. Organisation should develop incident response plans for different<br>kinds of scenarios including small incidents like malware infections all the way to big incidents that require system restoration.</p><blockquote>Checklist for Aspect 6:</blockquote><blockquote>✓ Incident response plans (including different kinds of security incidents) are<br>developed according to different scenarios.</blockquote><blockquote>✓ Systems and data are backup regularly, the backups are taken<br>offline (and even offsite).</blockquote><blockquote>✓ Restore procedures are drilled to make sure that the backup can be restored properly.</blockquote><h3>Aspect 7: User Awareness</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/768/1*lDsJjKN5rVr65K7XjRCxYQ.jpeg" /></figure><p>Users are the weakest links in cyber security. 95% security incidents involved human as a contributing factor. Organisations should ensure that staff understand their roles and responsibility in protecting information assets of the organisation.</p><blockquote>Checklist for Aspect 7:</blockquote><blockquote>✓ Staff should be reminded their roles and responsibility in protecting information assets of the organisation regularly, e.g. by staff awareness training.</blockquote><blockquote>✓ Drills (e.g. simulated phishing attacks) can be performed to test the readiness of staff against common cyber attack.</blockquote><p><strong>Feel free to contact me for any further question or consulting services.</strong></p><p><strong>Architect Yeung, 26 January 2022</strong></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=aba94bb9b5f" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Differences between Machine Learning and Classical Programming]]></title>
            <link>https://architectyeung.medium.com/jdifference-between-machine-learning-and-classical-programming-7b2b534b71ff?source=rss-3f56e5f686f5------2</link>
            <guid isPermaLink="false">https://medium.com/p/7b2b534b71ff</guid>
            <category><![CDATA[technology]]></category>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <dc:creator><![CDATA[Architect Yeung]]></dc:creator>
            <pubDate>Fri, 14 Jan 2022 06:26:36 GMT</pubDate>
            <atom:updated>2023-12-31T22:16:32.281Z</atom:updated>
            <content:encoded><![CDATA[<p><em>This article explains the key difference between Machine Learning and Classical Programming. Also, it highlights THREE different types of Machine Learning to give you some ideas and high-level understanding about ML.</em></p><p>Machine learning involves teaching a computer to recognise patterns by example, rather than programming it with specific rules. These patterns can be found in the data. In other words, ML is about creating algorithms (or a set of rules) that learn from complex functions (patterns) from data and make predictions on it (a form of “narrow AI”). ML learns from data and can be reused for unseen, future, or new data without rewriting code. Put another way, with ML, you start with a problem, identify data associated with that problem, use an algorithm to then model that problem, and generate output.</p><blockquote>Classical Programming uses data and static, hard-coded rules to generate results whereas Machine Learning uses historical data or output, combined with new data, to generate new rules in a training model.</blockquote><h3>Three Types of Machine Learning</h3><ol><li>Supervised Learning</li><li>Unsupervised Learning</li><li>Reinforcement Learning</li></ol><h3>Supervised Learning</h3><p>Supervised learning is a method where a model learns from a data set containing input values and paired output values that you would like to predict.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*j79YG5ngyDkAf1rJo4mZ9Q.png" /></figure><h3>Unsupervised Learning</h3><p>Unsupervised learning is a method in which the training model learns from data without any guidance. The objective is pattern and structure recognition.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*WOWQJZfBJjpvk-6u7V8oOg.png" /></figure><h3>Reinforcement Learning</h3><p>Reinforcement learning is a method in which the training model learns from its environment by being rewarded for correct moves and punished for incorrect moves.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ZejWgVdBtbq6WNvagbK0OQ.png" /></figure><p>Other reading: <a href="https://architectyeung.medium.com/machine-learning-basics-2b0aec360fa7"><strong>Machine Learning Basics</strong></a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=7b2b534b71ff" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Machine Learning Basics]]></title>
            <link>https://architectyeung.medium.com/machine-learning-basics-2b0aec360fa7?source=rss-3f56e5f686f5------2</link>
            <guid isPermaLink="false">https://medium.com/p/2b0aec360fa7</guid>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[technology]]></category>
            <dc:creator><![CDATA[Architect Yeung]]></dc:creator>
            <pubDate>Thu, 13 Jan 2022 13:12:20 GMT</pubDate>
            <atom:updated>2022-01-13T13:12:20.599Z</atom:updated>
            <content:encoded><![CDATA[<h3><strong>What is Machine Learning?</strong></h3><p>Machine learning (ML) is the process of training computers, using math and statistical processes, to find and recognise patterns in data. After patterns are found, ML generates and updates training models to make increasingly accurate predictions and inferences about future outcomes based on historical and new data. For example, ML could help determine the likelihood of a customer purchasing a particular product based on previous purchases by the user or the product’s past sales history.</p><p>Building ML applications is an iterative process that involves a sequence of steps. To build an ML application, follow these general steps:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*kLZqac8BAlebUNl7_aBabw.png" /></figure><h3>Key Terms in Machine Learning</h3><p><strong>ML Model: </strong><br>The output of an ML algorithm trained on a data set used for data prediction.</p><p><strong>Training: </strong><br>The act of creating a model from historical data called Testing Dataset.</p><p><strong>Testing:</strong><br>Measuring the performance of a model on Testing Dataset.</p><p><strong>Deployment:</strong><br>Integrating a ML model into a production pipeline or application.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*y8qbWfHMWxLyf1O8DoGrTA.png" /></figure><h3>Machine Learning Flywheel</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/886/1*HLcuIUR0RUUqP65Wf7FP_g.png" /></figure><p>ML flywheel uses data collected from parts of a business operation, uses a model to predict future outcomes, and provides ways to continuously improve efficiency and develop new operational capabilities and business practices. With ML, increasing predictions improve growth and efficiency. This leads to more usage and data, completing the feedback loop and reinforcing all parts of the flywheel.</p><h3>How can machine learning help me?</h3><p>Machine learning can continuously improve results, which means training models can become a part of almost any decision-making process. Machine learning can ingest limitless amounts of data, produce timely analysis and assessment, identify trends and patterns, and generate predictive forecasts.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*0YfBF6cGq1FUu-ReNbypBA.png" /></figure><h3><strong>Some examples of Machine Learning being used today</strong></h3><p>Example 1: Ride-sharing apps like Uber and Lyft use data to lower wait times, predict demand, and optimize price setting.</p><p>Example 2:Online shopping sites use ML to customize search results and improve product recommendations. Financial institutions use AI to recognize content on mobile check deposits.</p><p>Example 3: Credit- or debit-card transaction businesses use ML to scan for fraud. Research suggests that more than 8 billion digital voice assistants will be powered by AI and ML over the next few years.</p><p>Others: Industrial companies use AI and ML services for asset management. This includes using computer vision for equipment monitoring and defect detection, or analyzing operational machine behavior data to enable predictive maintenance. Customer service organizations use ML to transcribe and analyze live and archived calls for sentiment scores. ML can also help prioritize based on categorized customer feedback, and enable software to provide agents with answers to questions as they are being asked.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=2b0aec360fa7" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Five Principles of DevOps]]></title>
            <link>https://architectyeung.medium.com/five-principles-of-devops-bd20f4e4eac5?source=rss-3f56e5f686f5------2</link>
            <guid isPermaLink="false">https://medium.com/p/bd20f4e4eac5</guid>
            <category><![CDATA[agile-development]]></category>
            <category><![CDATA[cloud-architecture]]></category>
            <category><![CDATA[devops]]></category>
            <category><![CDATA[ci-cd-pipeline]]></category>
            <category><![CDATA[cloud-computing]]></category>
            <dc:creator><![CDATA[Architect Yeung]]></dc:creator>
            <pubDate>Tue, 11 Jan 2022 15:21:17 GMT</pubDate>
            <atom:updated>2022-01-11T15:21:17.780Z</atom:updated>
            <content:encoded><![CDATA[<p>This article summaries some core concepts of applying DevOps. AWS technologies and experiences would be taken as examples, to illustrate these concepts. Let’s start with the following <strong>five</strong> principles:</p><p><em>№1: Performing Frequent but Small Updates</em></p><p><em>№2: Adopting Microservice Architecture</em></p><p><em>№3: Infrastructure Automation</em></p><p><em>№4: Continuous Integration and Continuous Delivery</em></p><p><em>№5: Logging and Monitoring</em></p><h3><strong>Principle I: Performing Frequent but Small Updates</strong></h3><p>One fundamental practice is to perform very frequent but small updates. This is how organizations innovate faster for their customers. These updates are usually more incremental in nature than the occasional updates performed under traditional release practices.</p><p>Frequent but small updates make each deployment less risky. They help teams address bugs faster because teams can identify the last deployment that caused the error. Although the size of updates will vary, organizations using a DevOps model deploy updates much more often than organizations using traditional software development practices.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/764/1*sA68CSD6xDEY6RgihxKKZg.png" /></figure><h3><strong>Principle II: Adopting Microservice Architecture</strong></h3><p>What’s Microservice Architecture?</p><p>The microservice architecture decouples large, complex systems into simple, independent projects. Applications are broken into many individual components (services) with each service scoped to a single purpose or function and operated independently of its peer services and the application as a whole.</p><p>Why Microservice Architecture?</p><p>Organizations might also use a microservices architecture to make their applications more flexible and enable quicker innovation. This architecture reduces the coordination overhead of updating applications, and when each service is paired with small, agile teams who take ownership of each service, organizations can move more quickly.</p><p>The traditional development cycle always takes longer time to deploy new updates because project teams need to make sure the integrations between various components working well with the new update in the complex “monolith” application. The situation becomes even worse when the application involves more team collaboration and deploys more features / integrations.</p><p>As a result, people start thinking a new collaboration model in terms of architecture and team size. The concepts of “2-pizza team” and microservice architecture are introduced in the recent market. Each “2-pizza team” involves 5 to 7 team members, focusing on building their own microservices which work as web services / APIs. Also, each microservice works independently for a single function. The project team can make a combination of these microservices (i.e. building blocks) to build the whole application.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/850/1*hkSqleS_4o_57bY30s8MUQ.png" /></figure><h3><strong>Principle III: Infrastructure Automation</strong></h3><p>However, there are some drawbacks of the combination of microservices and increased release frequency. It leads to significantly more deployments which can present operational challenges and risks.</p><p>On one hand, we like to increase the development agility; on the other hand, we need to implement more on Management, Control and Automation. Thus, DevOps practices like Continuous Integration CI and Continuous Delivery CD try to solve these issues and let organizations deliver rapidly in a relatively safe and reliable manner.</p><p>Furthermore, infrastructure automation practices, like Infrastructure as Code and configuration management, help to keep computing resources elastic and responsive to frequent changes.</p><p>In addition, the use of monitoring and logging helps engineers track the performance of applications and infrastructure so they can react quickly to problems.</p><p>Benefits of leveraging Infrastructure as Code:</p><p>· Scalability (anything manual is not scalable)<br>· Reliability<br>· Reproduction / Duplication<br>· Environment Consistency<br>· Auditability / Record Keeping<br>· Security<br>· Governance</p><p>Tools for Configuration Management:</p><p>· AWS CloudFormation<br>· Ansible<br>· Saltstack<br>· Chef<br>· Puppet</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/850/1*xnwA4eMS5AYZRFZJ5JC62A.png" /></figure><h3><strong>Principle IV: Continuous Integration CI and Continuous Delivery CD</strong></h3><p>Benefits of adopting CI and CD:</p><p>· Help to prove code quality and function repeatedly with predefined results.</p><p>· Enable monitoring, testing and validation.</p><p>· Many plugins are available in the market for integrating existing processes.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/850/1*iEuTqAEgaogw725M28JMIA.png" /></figure><h3><strong>Principle V: Logging and Monitoring</strong></h3><p>· Maintain visibility and audibility of activity in your application infrastructure.</p><p>· Assess how application and infrastructure performance impact end-user experience.</p><p>· Gain insight into the root causes of problems or unexpected changes.</p><p>· Support services that must be available 24/7 as a result of continuous integration/continuous delivery.</p><p>· Create alerts based on thresholds you define.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/850/1*RZVc_MVah6FEIRatnWqGrw.png" /></figure><p>For any question, contact me at <a href="/service/mailto:architectyeung@gmail.com">architectyeung@gmail.com</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=bd20f4e4eac5" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>