<?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 Javier Toledo on Medium]]></title>
        <description><![CDATA[Stories by Javier Toledo on Medium]]></description>
        <link>https://medium.com/@javier_toledo?source=rss-72b6853a68ea------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*uQA6Ub_6Z3qlMHsE_IfYRQ.jpeg</url>
            <title>Stories by Javier Toledo on Medium</title>
            <link>https://medium.com/@javier_toledo?source=rss-72b6853a68ea------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Tue, 17 Nov 2026 10:26:20 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@javier_toledo/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[LLMs and Embeddings 101: Unleash the Power of GPT-4 with Unbounded Long-term Memory]]></title>
            <link>https://medium.com/the-theam-journey/llms-and-embeddings-101-unleash-the-power-of-gpt-4-with-unbounded-long-term-memory-edd77b83e536?source=rss-72b6853a68ea------2</link>
            <guid isPermaLink="false">https://medium.com/p/edd77b83e536</guid>
            <category><![CDATA[vector-database]]></category>
            <category><![CDATA[embedding]]></category>
            <category><![CDATA[llm]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[embeddings-search]]></category>
            <dc:creator><![CDATA[Javier Toledo]]></dc:creator>
            <pubDate>Thu, 29 Jun 2023 11:39:38 GMT</pubDate>
            <atom:updated>2023-07-03T16:17:25.184Z</atom:updated>
            <content:encoded><![CDATA[<h4>An introductory-level catch-up for anyone interested in going beyond ChatGPT</h4><figure><img alt="Midjourney. Prompt: a galaxy of embeddings inside of a robot mind" src="https://cdn-images-1.medium.com/max/1024/1*-QbkfF6ddzLQssLSyjHXzA.png" /><figcaption>Generated with Midjourney. Promt: <em>a galaxy of embeddings inside of a robot mind</em>.</figcaption></figure><p>In the dynamic world of Artificial Intelligence, the tools and concepts we use are continually evolving. Language models, an integral part of this landscape, have grown in complexity and capability, resulting in the creation of Large Language Models (LLMs) such as GPT-4. Alongside LLMs, the concept of semantic embeddings has emerged, unlocking new ways to represent and utilize information. This article serves as a guide, detailing the essentials of LLMs and embeddings, and showcasing how they can be leveraged to extend the power of AI models, providing them with a form of unbounded long-term memory.</p><h3>What is a Large Language Model (LLM)?</h3><p>A Large Language Model is a type of AI model that is designed to generate human-like text. These models are trained on vast amounts of text data, enabling them to predict the next word in a sentence, given the context of the previous words. The “largeness” in LLMs refers to both the training data’s size and the model’s complexity, which often involves billions, or even trillions, of parameters (Each of the individual connections inside of the model). GPT-4, for example, leverages transformer-based architectures to process and generate text in a contextually and semantically rich way.</p><p>Thanks to the expansive and high-quality training data they employ, these models have acquired a certain level of problem-solving and reasoning capabilities. These skills are inherited from the vast array of language and literature they’ve been trained on. Consequently, they offer innovative ways to process and comprehend documents and data. Moreover, they provide more user-friendly and intuitive interfaces for human interaction.</p><h3>What does mean to “train” a model?</h3><p>Training a model, such as GPT-4, is an intricate and resource-intensive process. Drawing inspiration from the workings of biological brains, a large AI model is a network comprised of millions of nodes (or artificial neurons), connected by billions of weighted links. The training process necessitates the calculation of a specific set of weights to make the neural network function in the desired manner.</p><p>To illustrate, the fundamental objective of an LLM like GPT-4 is, quite simply, to predict the most sensible next word given a piece of text. To accomplish this, OpenAI engineers utilized a substantial corpus of text examples, presenting each document to the network and systematically adjusting the weights until the correct words began to surface.</p><p>Training a large language model is a remarkably complex undertaking, necessitating months of computation on specialized hardware and finely honed algorithms. Hence, it’s not a venture that just any organization can embark upon. To harness the power of LLMs using your own data — a prevalent use case — it’s more common not to train a custom model, but to utilize an existing model coupled with a database of embedding-indexed data.</p><h3>What are the Embeddings?</h3><p>Embeddings are numerical representations that encapsulate the semantic meanings and relationships among words, paragraphs, or even entire documents. While that may sound complex, consider this analogy: The word “apple” is semantically closer to the word “fruit”, signifying that an apple is a type of fruit, a relationship our brains naturally discern. However, we view the word “apple” as semantically distant from the word “car”. Embeddings provide a mathematical depiction of these meanings, maintaining the relative semantic distances as we cognitively perceive them. But how? By representing them as vectors or points in high-dimensional space. In this context, “apple” would be positioned closer to “fruit” and further from “car”, thereby reflecting the semantic relationships among these words. Specialized models are used to generate these embeddings.</p><p>In mathematical terms, a vector can be understood as a collection of coordinates that represent a point in space, corresponding to the size or dimensions of the vector itself. For instance, a vector of three numbers could denote a specific point in space, such as latitude, longitude, and altitude. Embeddings function on a similar principle, but typically involve thousands of dimensions, thus creating a high-dimensional space. Even though we cannot visually represent these in a way that’s comprehensible to humans, we can still manipulate them mathematically. For example, to calculate the distance between two text pieces, which would be a numerical measurement of how much their semantics are related.</p><h3>Generating Embeddings for custom data</h3><p>Embeddings can be created with specialized AI models that are specifically designed for this. OpenAI provides its own API endpoint to generate embeddings, but there are other models that can be used, the important thing is that you use the same model for all the documents you’re handling. The embeddings model processes a given text and outputs the corresponding high-dimensional vector.</p><p>This operation is not reversible, given an embedding, you can’t get the original text again, so the original text must be kept in order to retrieve it later. The interest in generating embeddings is that once we have them for two documents, we can compare how close or far they are from each other from a semantics point of view.</p><p>To enhance the efficiency of our embeddings and circumvent the size constraints inherent in embedding models, it might be necessary to partition large documents into smaller segments. The more distinct these segments are, the better. For instance, if we were to generate embeddings for this article, we would likely create one for each individual section.</p><h3>Specialized Databases for Embeddings</h3><p>When dealing with extensive text collections in Language Learning Models (LLMs), the large embeddings created require a special operation for retrieval, known as Nearest Neighbours Search. This operation is not supported by most conventional databases, necessitating the use of specialized databases adept at handling high-dimensional data.</p><p>There are many solutions designed for efficient storage, search, and management of vectors like <a href="https://ai.facebook.com/tools/faiss/">Faiss by Meta</a>, <a href="https://milvus.io/">Milvus</a>, <a href="https://qdrant.tech/">Qdrant</a>, <a href="https://www.trychroma.com/">Chroma DB</a>, or <a href="https://www.pinecone.io/">Pinecone</a>, among others​.</p><p>For more traditional setups, it is possible to use <a href="https://github.com/pgvector/pgvector">PostgreSQL with the pgvector extension</a>, which could be convenient if you’re planning to keep all the data in the same place​, or use the advanced vector search features of indexing engines as <a href="https://www.elastic.co/">ElasticSearch</a> or <a href="https://opensearch.org/">OpenSearch</a>.</p><h3>Using Embeddings for Semantic Search</h3><p>Once we have generated embeddings for all our documents and inserted them in a database that supports vector searches, they can be used to perform semantic searches, a type of search that understands the searcher’s intent and the contextual meaning of terms to generate more relevant results.</p><p>Because embeddings capture the semantic meaning of words and sentences, they can be used to find the “closest” (in terms of vector distance) entries in a database to a given query or question.</p><p>For example, if we input a query like “juicy fruit”, a semantic search powered by embeddings might return entries containing “apple” or “orange”, even if the exact phrase “juicy fruit” is not present. This is because the embeddings for “juicy fruit” and “apple” are close in the high-dimensional space.</p><h3>In-Context Learning and Extending the LLM’s Long-Term Memory</h3><p>A key aspect of LLMs like GPT-4 is their ability to generate text that is contextually relevant. However, these models traditionally do not have long-term memory, meaning they don’t remember information beyond the current context window (a few thousand tokens for GPT-4).</p><p>To expand this capacity and give the model a form of “long-term memory,” we can perform semantic searches for any user input, select the closest documents, and inject them as part of the model prompt to augment its knowledge of the matter before answering. This technique is known as in-context learning, because instead of training the model with new data, we inject relevant information as part of the prompt within the model’s context window.</p><p>For instance, if we’ve previously generated an embedding for a phrase like “Paris is the capital of France”, we can feed this embedding into the model when we ask it about French landmarks. Even though “Paris is the capital of France” is not in the immediate context, the model will still have this information available thanks to the injected embedding.</p><p>This method enables the model to recall and use information beyond its standard context window. This approach is particularly useful in applications where the model needs to maintain consistency or recall information over a large set of documents like a company’s Wiki or the documentation of a technical tool.</p><h3>Conclusion</h3><p>The development of LLMs like GPT-4 has been a major stride in AI and NLP with important implications at many levels, but when organizations want to use them with their own data, training a dedicated LLM from scratch is probably not the best solution. With the power of embeddings, we can store and use semantic information in previously unfeasible ways, allowing us to “teach” commercially available LLMs to work with data outside of their training corpus without having to re-train them.</p><p>This technique enables the possibility to any organization to build specialized LLM-powered tools that take advantage of private or very specific or private data and knowledge.</p><p>Do you want to know more? Visit our <a href="https://building.ellmental.com/">‘Open Kitchen’ page</a> to find everything we’ve learned on our AI journey. Don’t hesitate to contact us if you need help!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=edd77b83e536" width="1" height="1" alt=""><hr><p><a href="https://medium.com/the-theam-journey/llms-and-embeddings-101-unleash-the-power-of-gpt-4-with-unbounded-long-term-memory-edd77b83e536">LLMs and Embeddings 101: Unleash the Power of GPT-4 with Unbounded Long-term Memory</a> was originally published in <a href="https://medium.com/the-theam-journey">The Agile Monkeys’ Journey</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[LLMs Raise the Bar in Software Development: Adapting to a New Era of Democratized 10x Productivity]]></title>
            <link>https://medium.com/the-theam-journey/llms-raise-the-bar-in-software-development-adapting-to-a-new-era-of-democratized-10x-productivity-fa448a702593?source=rss-72b6853a68ea------2</link>
            <guid isPermaLink="false">https://medium.com/p/fa448a702593</guid>
            <category><![CDATA[chatgpt]]></category>
            <category><![CDATA[github-copilot]]></category>
            <category><![CDATA[software-development]]></category>
            <category><![CDATA[productivity]]></category>
            <category><![CDATA[ai]]></category>
            <dc:creator><![CDATA[Javier Toledo]]></dc:creator>
            <pubDate>Tue, 09 May 2023 21:51:46 GMT</pubDate>
            <atom:updated>2023-05-09T21:51:46.278Z</atom:updated>
            <content:encoded><![CDATA[<h4><a href="https://en.wikipedia.org/wiki/Large_language_model"><strong>Large Language Models (LLMs)</strong></a><strong> like </strong><a href="https://openai.com/product/gpt-4"><strong>GPT-4</strong></a><strong> are transforming how we work and learn in the software development industry. With cutting-edge knowledge now available to everyone, how will developers thrive in a post-GPT world?</strong></h4><figure><img alt="A software developer and a futuristic robot writing code together during a pairing session" src="https://cdn-images-1.medium.com/max/1024/1*-2AtmkW23InrSTspIRbTnQ.png" /></figure><p><em>This is a personal opinion piece with my current view on what’s coming. I might be plain wrong or change opinions tomorrow, so I invite you to respond to this post and share your thoughts and perspectives on this topic too.</em></p><h3>Introduction</h3><p><a href="https://en.wikipedia.org/wiki/Large_language_model">Large Language Models (LLMs)</a> such as <a href="https://openai.com/product/gpt-4">OpenAI’s GPT-4</a> are revolutionizing the software development industry by democratizing access to advanced knowledge and programming expertise. <a href="https://github.com/features/copilot">GitHub Copilot</a>, for example, brings LLM capabilities directly into developers’ Integrated Development Environments (IDEs), and the upcoming release of <a href="https://github.com/features/preview/copilot-x">GitHub Copilot X</a> promises even more advanced features.</p><p>These tools will make us faster and produce higher-quality code, thus producing below-average code will soon become unacceptable, and developers must adapt to stay competitive.</p><h3>The Evolving Role of the AI-enhanced Software Developers</h3><p>In this new landscape, the day-to-day work of a software developer could shift towards a role more similar to that of a lead developer, with less emphasis on writing code and more focus on ensuring quality, reviewing code, deeply understanding business requirements, and agreeing with their team on code style, architecture, and boundaries. One of the most critical skills that software developers might need to acquire soon is their ability to build prompts and evaluate, fix, accept, or reject code and documentation <a href="https://githubnext.com/projects/copilot-for-pull-requests/#resolving-issues-with-ai">proposed by LLMs</a>.</p><p>Additionally, LLMs can significantly accelerate the learning process for developers. Previously, learning a new tool or language could take weeks or months of practice, with developers often becoming stuck in the process. The learning experience typically involved long periods of confusion followed by small moments of realization. <a href="https://docs.boosterframework.com/">LLMs can absorb the documentation of any tool</a> and, with <a href="https://www.howtogeek.com/881159/chatgpt-can-now-surf-the-web-with-plugins/">browsing capabilities</a>, inspect the entire internet to find solutions to blockages in seconds. This drives developers through a series of realization moments, one after another, without getting stuck in obstacles. As a result, there’s no reason not to produce not only good but also idiomatic code for any technology you can imagine.</p><blockquote>The post-GPT software engineers will barely write code themselves but still hold responsibility for any code pushed.</blockquote><p>As I see it, the software engineer of the post-GPT world will need more knowledge of quality standards, design patterns and architecture, security, understanding of business processes, and soft skills. All of them already desirable qualities for excellent developers, but now made essential, because knowing how to write code for a specific stack won’t be enough anymore. They will need to master transferable programming skills that make them effectively polyglot, being able to use any programming language or framework in no time, and there will be no excuses for undocumented or poorly tested code.</p><h3>Thriving in a Post-GPT Software Development World</h3><p>Many developers are afraid of the upcoming changes, and there are good reasons to be concerned. We come from a world with a massive demand for anyone who could contribute to creating software. Knowing how to code could easily get you a great job offer with great perks. Now we are headed to a world where one developer with the proper set of skills could produce as much work as an entire team does today.</p><p>To remain competitive in such a landscape, developers must rethink their strategies and adapt to the new situation. Embracing continuous learning, specialization into hard-to-master topics, focusing on creative problem-solving, honing soft skills, and staying up-to-date with industry trends and emerging technologies will be essential. By cultivating their ability to evaluate others’ code and effectively guide its improvement critically, developers can navigate the evolving demands of the job market and thrive in a post-GPT software development world.</p><blockquote>I think the world is gonna find out that if you can have ten times as much code at the same price, you can just use even more — <a href="https://youtu.be/L_Guz73e6fw?t=5812">Sam Altman in the Lex Fridman podcast (1:36:52)</a></blockquote><p>But is it worth the effort? Many are worried that having AI-enhanced developers who do 10x the work will result in 10x fewer developers needed. I believe that at least during the next few years, even if productivity is increased by 10x, the amount of work required and what we expect from software might easily outgrow the productivity increase. AI is not here to compete against competent software developers but to allow us to work on all those ideas that never left the backlog because we had to focus on delivering the ones from the top.</p><p><em>As a personal opinion piece, I welcome your thoughts, experiences, and perspectives on this topic. Please feel free to share them and contribute to the conversation!</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=fa448a702593" width="1" height="1" alt=""><hr><p><a href="https://medium.com/the-theam-journey/llms-raise-the-bar-in-software-development-adapting-to-a-new-era-of-democratized-10x-productivity-fa448a702593">LLMs Raise the Bar in Software Development: Adapting to a New Era of Democratized 10x Productivity</a> was originally published in <a href="https://medium.com/the-theam-journey">The Agile Monkeys’ Journey</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[You can no longer afford time amnesia in your software systems.]]></title>
            <link>https://medium.com/the-theam-journey/you-can-no-longer-afford-time-amnesia-in-your-software-systems-52304e79d1d3?source=rss-72b6853a68ea------2</link>
            <guid isPermaLink="false">https://medium.com/p/52304e79d1d3</guid>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[event-driven-systems]]></category>
            <category><![CDATA[event-sourcing]]></category>
            <category><![CDATA[machine-learning]]></category>
            <dc:creator><![CDATA[Javier Toledo]]></dc:creator>
            <pubDate>Tue, 14 Mar 2023 19:26:27 GMT</pubDate>
            <atom:updated>2023-03-15T09:48:43.612Z</atom:updated>
            <content:encoded><![CDATA[<h3>In the AI-driven world, you can no longer afford time amnesia in your software systems.</h3><h4>Event-driven machine learning will enable a new generation of businesses that will be able to make incredibly thoughtful decisions faster than ever, but is your data ready to take advantage of it?</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*BMDn-3fXEP4uZadTq6f55A.png" /></figure><p>When you look at the speed of innovation in the AI world, it seems like each business out there will eventually fall within one of these two categories: <a href="https://hbr.org/2020/01/competing-in-the-age-of-ai">the ones that took advantage of AI and the ones that went out of business because they could no longer compete</a>. And <strong>the key differentiation factor is going to be the amount and quality of their data</strong>. Here, let me especially highlight the word <strong>quality</strong> because <strong>knowing what, when, and why something happened in your company will be crucial</strong>.</p><p>But <strong>how can you be certain that you can recover this information since you turned your system on for the first time</strong>? This is exactly the core value of <a href="https://martinfowler.com/eaaDev/EventSourcing.html">event sourcing</a>: Instead of storing only the current state of a system, <strong>everything that happens is stored permanently</strong> as a long series of immutable events. Then, <strong>the state shown to the users is calculated on demand, but the events are never discarded</strong>, they can always be inspected, reprocessed, or forwarded to other places.</p><p><a href="https://event-driven.io/en/when_not_to_use_event_sourcing/"><strong>Not all developers are willing to adopt this way of storing data</strong> <strong>though</strong></a>. <a href="https://martinfowler.com/eaaDev/EventSourcing.html">Event sourcing</a> and <a href="https://en.wikipedia.org/wiki/Command%E2%80%93query_separation#Command_Query_Responsibility_Separation">CQRS</a>, an intimately related pattern, require a slightly different way of thinking. Software developers that are not used to them tend to fall back into implementing <a href="https://en.wikipedia.org/wiki/Representational_state_transfer">RESTful solutions</a> instead, as they’re <a href="https://eventmodeling.org/"><strong><em>perceived</em></strong></a> simpler, but at design time <a href="https://www.forbes.com/sites/forbestechcouncil/2022/06/27/data-culture-what-it-is-and-how-to-make-it-work/?sh=1557ffba2096">it’s hard to predict who and when will need the extra data</a>, so why going the extra mile? Why making the extra investment on development time and data storage?</p><blockquote>In the era of AI, a company can no longer afford to discard a whole dimension of their data, especially when it is as relevant as time.</blockquote><p>That is exactly the risk, once you’ve discarded the time-series data, it becomes unrecoverable. <strong>We’re sacrificing a dimension of the data that could become of high relevance in the mid-term for a perceived short-term benefit</strong>.</p><p>Data science teams mitigate this problem using <a href="https://en.wikipedia.org/wiki/Change_data_capture">Change Data Capture (CDC)</a>. Most databases keep an operations log that provide a reliable source of differential changes. But <strong>CDC misses another important variable: the semantics of each change</strong>, that is, why the change happened. It’s not the same to say “the order has been updated” as saying “the order has been paid,” the second statement is way more relevant from a business analytics point of view.</p><p>A development team that is aware of the importance of data will always make an effort to generate enough metadata to infer the context. Still, <strong>humans will make mistakes, and more frequently, we will easily decide to deprioritize this when there’s a hard deadline</strong>. A team that uses <strong>event sourcing generates high-quality data by design, </strong>without the team needing to make a conscious decision or extra effort.</p><blockquote>Event sourcing removes the responsibility from the development team to decide whether a change must be stored, because the only way to change the system state is by creating an event that encapsulates what changed, when, and why.</blockquote><p>But the advantages of storing events don’t end there. Events are also structurally different, <strong>they can be easily sorted and consumed in progressive chunks</strong>, and as they’re immutable, <strong>there’s no risk that a piece of data you’ve already processed gets out of sync because other process updates it</strong>. All event consumers can work assuming that they only need to process the next event in the queue and won’t miss a byte of data. Hence, <strong>they enable, </strong><a href="https://medium.com/swlh/the-engineers-guide-to-event-driven-architectures-benefits-and-challenges-3e96ded8568b"><strong>among other things</strong></a><strong>, what some authors are starting to describe as </strong><a href="https://medium.com/phi-skills/event-driven-machine-learning-52680a97cf64"><strong>event-driven machine learning</strong></a>:</p><ul><li><strong>Collecting data becomes easier</strong> with a real-time flow of events available to data analysts, as opposed to relying on static data dumps that provide a snapshot of a database at a particular moment. The event stream is akin to a live sports broadcast, which <strong>can be paused to view specific details in a scene, replayed to analyze interesting moments from the past, or monitored in real-time to observe the most current data</strong>.</li><li><a href="https://towardsdatascience.com/how-to-apply-continual-learning-to-your-machine-learning-models-4754adcd7f7f"><strong>Continual learning</strong></a> becomes possible, this is a technique that <strong>emulates the human ability to adapt their behavior in a dynamic environment.</strong> The models are re-trained with recent events to continuously adapt them to changing environments (i.e. <a href="https://www.invisibly.com/learn-blog/netflix-recommendation-algorithm/">the recommendation system of Netflix requires constant fine-tuning in response to the introduction of new shows and evolving viewer preferences</a>).</li><li>Fine-grained data manipulation enables data-driven decision-making. <strong>Analysts can view data from different angles by reprocessing events and uncover previously inaccessible information</strong>. Each event represents <strong>a change that can be analyzed at an individual level</strong>, such as individual transactions in a bank account, <strong>or aggregated to see a broader picture at a particular time</strong>, such as the account balance. <strong>Data can also be processed to create new perspectives</strong>, like categorizing expenses by type.</li></ul><p>With nowadays tools, using the data in this way can be expensive, and still require the expertise of specialized developers or data scientists, but <strong>we’re expecting advancements in large language models capabilities to provide natural language interfaces for manipulating and visualizing data.</strong> These new data manipulation models will make these techniques available for anyone in an organization, fostering actual <a href="https://www.gartner.com/smarterwithgartner/the-key-to-establishing-a-data-driven-culture#:~:text=A%20key%20characteristic%20of%20a%20data-driven%20culture%20is,transparent%20about%20data%20access%20restrictions%20and%20governance%20methods.">data-driven business cultures</a>.</p><blockquote><strong>New advancements in large language models capabilities will provide natural language interfaces for manipulating and visualizing data, and make data-driven decisions at every level in an organization.</strong></blockquote><p>In short, with all the amazing AI products we’ve seen recently, such as DALL·E or ChatGPT, <strong>we’re barely scratching the surface of what’s to come</strong>. It’s clear that AI is here to stay and <strong>will deeply affect how we conduct business</strong>. To stay competitive in this new AI-driven world, <strong>companies will need to collect high-quality data of everything that happens in their systems</strong>.</p><p><strong>Event sourcing captures high-quality data by design</strong>, packing the what, why, and when for every change, recording events that can be compared to the individual frames of a high-definition security camera for your data.</p><p>This constant stream of events provides extra benefits. <strong>Data can be analyzed or manipulated in real-time</strong>, and <strong>previous states can be replayed</strong>. <strong>Event-driven machine learning models enable accurate predictions and automatic reactions to complex scenarios</strong>, working as a restless security guard that continuously monitor the cameras and find inefficiencies and risks.</p><p>For all these reasons, we believe that <strong>companies that adopt event-driven designs and learn to take advantage of event-driven machine learning models will have a strong competitive advantage and lead the market in this new AI-driven generation</strong>.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=52304e79d1d3" width="1" height="1" alt=""><hr><p><a href="https://medium.com/the-theam-journey/you-can-no-longer-afford-time-amnesia-in-your-software-systems-52304e79d1d3">You can no longer afford time amnesia in your software systems.</a> was originally published in <a href="https://medium.com/the-theam-journey">The Agile Monkeys’ Journey</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Making and escalating decisions as a distributed team]]></title>
            <link>https://medium.com/the-theam-journey/making-and-escalating-decisions-as-a-distributed-team-762e6769db68?source=rss-72b6853a68ea------2</link>
            <guid isPermaLink="false">https://medium.com/p/762e6769db68</guid>
            <category><![CDATA[distributed-teams]]></category>
            <category><![CDATA[remote-working]]></category>
            <category><![CDATA[decision-making]]></category>
            <category><![CDATA[leadership]]></category>
            <dc:creator><![CDATA[Javier Toledo]]></dc:creator>
            <pubDate>Mon, 01 Mar 2021 20:47:45 GMT</pubDate>
            <atom:updated>2021-03-01T20:48:22.044Z</atom:updated>
            <content:encoded><![CDATA[<h4>Learn to differentiate between one-way and two-way doors.</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*tZlHpYK8sKvpS3Rjn7ajGg.jpeg" /><figcaption>Photo by <a href="https://unsplash.com/@randomlies?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Ashim D’Silva</a> on <a href="https://unsplash.com/s/photos/wrong-door?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Unsplash</a></figcaption></figure><p>As a distributed and manager-lesser team, we want to make sure that each of us has the autonomy to make decisions and execute them without needing to ask the <strong><em>boss™</em></strong> for permission.</p><blockquote>Read <strong><em>boss™</em></strong> here as a manager, lead engineer, VP, CEO, CTO, the Oracle of Apollo at Delphi, $DEITY, your cat, a couple of dice, flipping a coin, or any other authority figure of any kind.</blockquote><p>But this is easier to tell than to put into practice. Some decisions should be taken with extra care and involve other team members and <strong><em>bosses™</em></strong> to ensure that they match the overall project and company strategy.</p><p>It’s hard, especially for the less experienced people, to differentiate when they can just execute an idea or escalate it. Lack of confidence can lead you to become too conservative and end up asking your <strong><em>boss™</em></strong> about the smallest decisions, introducing an extra overhead that will slow things down. On the contrary, if you lean towards the other side and act in a lone-ranger-reckless way, you can become a potential threat to your project’s continuity or even your company.</p><p>Jeff Bezos coined the terms “One-way” and “Two-way” doors in his<a href="https://www.sec.gov/Archives/edgar/data/1018724/000119312516530910/d168744dex991.htm"> AMZN 2015 Letter to shareholders</a>. He was explaining that to make Amazon become an “invention machine,” they would need to avoid “one-size-fits-all” (conventional) decision making and learn to differentiate between two kinds of decisions:</p><ul><li><strong>One-way doors</strong>: decisions that are consequential and nearly irreversible; should be made methodically, carefully, slowly, with great deliberation and consultation.</li><li><strong>Two-way doors</strong>: decisions that are changeable and reversible; should be made quickly by high judgment individuals or small groups. If you’ve made a suboptimal decision, you can reopen the door and go back through.</li></ul><p>It’s easier to decide if you should escalate a decision by thinking about whether it’s a one-way or a two-way door. If you can undo an action with no consequences, it’s better to execute it. Otherwise, it’s better to make sure you’re doing the right thing and check with your team members and leaders.</p><p>Most decisions are two-way doors, so following this simple framework will make your team faster and more confident.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=762e6769db68" width="1" height="1" alt=""><hr><p><a href="https://medium.com/the-theam-journey/making-and-escalating-decisions-as-a-distributed-team-762e6769db68">Making and escalating decisions as a distributed team</a> was originally published in <a href="https://medium.com/the-theam-journey">The Agile Monkeys’ Journey</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Riding the Elliot hype rollercoaster: and now what?]]></title>
            <link>https://medium.com/the-theam-journey/riding-the-elliot-hype-rollercoaster-and-now-what-94878c50560a?source=rss-72b6853a68ea------2</link>
            <guid isPermaLink="false">https://medium.com/p/94878c50560a</guid>
            <category><![CDATA[headless-commerce]]></category>
            <category><![CDATA[ecommerce]]></category>
            <category><![CDATA[entrepreneurship]]></category>
            <category><![CDATA[open-source]]></category>
            <category><![CDATA[startup]]></category>
            <dc:creator><![CDATA[Javier Toledo]]></dc:creator>
            <pubDate>Thu, 02 Jul 2020 09:37:06 GMT</pubDate>
            <atom:updated>2020-07-07T12:34:31.976Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*rVkHSxhUfiTel8d6oajhQw.jpeg" /><figcaption>Photo by <a href="https://unsplash.com/@askkell?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Andy Kelly</a> on <a href="https://unsplash.com/s/photos/cow?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Unsplash</a></figcaption></figure><blockquote><strong>Update:</strong> As of July 7th, 2020, the linked tweets were deleted. I kept the screenshots for reference.</blockquote><p>Since the first appearance of the <a href="https://elliot.store/">Elliot</a> commerce platform <a href="https://www.gcimagazine.com/marketstrends/channels/onlinetv/Mobile-Platform-Brings-Together-Convenience-Content-and-Commerce-510335011.html?ref=producthunt">last year</a> <strong>we’ve been following the project very closely with no little excitement</strong>.</p><p>In 2016 I wrote an article about what I considered <a href="https://medium.com/the-theam-journey/introducing-the-swing-project-e5ccc73a2c7f">“The future of e-commerce”</a>, sharing my vision of all <strong>the features that a modern e-commerce platform should offer</strong> from scratch to make developers and the business owner’s lives easier. During the last four years, we’ve learned a few more lessons and I’d add some more items to the list, like taking advantage of <a href="https://medium.com/the-theam-journey/serverless-for-business-people-dd477959e7c1">Serverless technologies</a> to <strong>remove the need for managing infrastructure</strong>. These ideas have been our beacon for many of the things we’ve done, like <strong>our very new open-source framework </strong><a href="https://booster.cloud/"><strong>Booster</strong></a><strong>!</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*81uvQedX1Ml7gDYTlskrSQ.png" /><figcaption>High-scale e-commerce is challenging, it inspired us to create new ways to write code!</figcaption></figure><p>The first time we saw Elliot’s proposal <strong>we perceived it as the breath of fresh air that the industry needed</strong>, checking most of the boxes of what we’ve been missing in e-commerce, like offering a <strong>fully-fledged headless API</strong> that allows total creative freedom to build amazing user experiences for our customers or supporting all sort of devices. At the same time, they brought an <strong>innovative win-win model</strong> that made a lot of sense (100% free if you don’t sell), or <a href="https://medium.com/greener-pastures-by-elliot/e-commerce-is-never-going-to-look-the-same-again-thanks-to-elliot-760816b76142">attacking <strong>major pain points like integrations with payment and shipping services</strong></a>! All of that served in a SaaS + Serverless package that makes it <strong>easy to start and easier to extend and scale</strong> without the need of dedicated ops teams.</p><p>For all these reasons, <strong>we couldn’t be more excited when we saw their plans of releasing a new version on June 18th</strong>, would this mean the end of fighting Magento or Shopify limitations to build custom UIs and experiences? Was this too good to be true?</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1010/1*xRtpQSwHvkluag2sIwOA_Q.png" /><figcaption><a href="https://twitter.com/hello_iamelliot/status/1260238304152358912">Link to the Tweet</a></figcaption></figure><p>Well, it turns out that some things didn’t go as expected and <strong>the situation escalated quickly to a totally unexpected outcome</strong>. Firstly, they had to roll back the release, apparently not a big deal, we’ve all been in similar situations:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1010/1*u4zMKFIFtnJ2dN80gZxDug.png" /><figcaption><a href="https://twitter.com/sir_gee_ohhhhh/status/1273674522953158659">Link to the Tweet</a></figcaption></figure><p>A couple of days later, they shared that their infrastructure was being DDoSed at release time, so it seemed totally reasonable to delay it, I think that anyone can understand that.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1010/1*8kTmO_hdOqKReiNegsC-Iw.png" /><figcaption><a href="https://twitter.com/sir_gee_ohhhhh/status/1274146757518163968">Link to the Tweet</a> | <a href="https://docs.google.com/document/d/14y7Yq_rcNq50Xaa5US7Iz-EOqt3hNprUOVmFfRWk5rs/edit">Link to the letter in Google Docs</a></figcaption></figure><p>But in a totally unexpected turn, <strong>their CEO resigned over twitter</strong>:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1010/1*x9unLBRSJSbb83BjHCi-5Q.png" /><figcaption><a href="https://twitter.com/sir_gee_ohhhhh/status/1276213923474214913">Link to the Tweet</a></figcaption></figure><p>And <strong>a few days later, the whole project is canceled</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1010/1*aV0x18EmXyIfaKp1zqs9Yw.png" /><figcaption><a href="https://twitter.com/hello_iamelliot/status/1278056349729988608">Link to Tweet</a> | <a href="https://docs.google.com/document/d/1V3Sfm2ZNLxFjNFSExKxGkjZ3BRVBVIaAdvrdf8S5W44/edit">Link to the letter in Google Docs</a></figcaption></figure><p>I don’t know the details of what happened, the story, as it has been publicly explained, doesn’t make much sense, so <strong>we’re missing important pieces of information</strong> here to blame anyone. What is clear is that <strong>the release failed in a way that they thought wasn’t easily recoverable</strong>. For me, <strong>their (very well) marketed value proposition had no loose ends</strong>, but that’s only one of the thousand pieces that have to fit in place for a business to be sustainable.</p><p>What I know is that <strong>this is bad news for the e-commerce industry</strong>. The <strong>technology in this space has been stuck for too long</strong>, and we really need an impulse of innovation.</p><p><strong>Anyway, we can learn something valuable</strong> out of this: <strong>their value proposition made a lot of sense!</strong> They tested a different approach, both from technology and business perspectives, removing several of the long-claimed pain points, and <strong>got a clear market fit</strong>, with a significant part of the community becoming excited, they’ve shown us that <strong>a future with fewer limitations and more possibilities is possible for e-commerce</strong>.</p><p>I don’t know you, but <strong>I really wanted this project to succeed</strong>, their marketing and energy were amazing and contagious, and <strong>many of us might be feeling that they’re leaving a void behind</strong>. That’s why our team at <a href="https://www.theagilemonkeys.com/">The Agile Monkeys</a> has decided to <strong>start </strong><a href="https://spectrum.chat/boostercloud/commerce/lets-explore-how-to-build-a-modern-event-driven-e-commerce-in-booster~f3755940-a87d-427c-8fc9-2196c34ec38d"><strong>an open-source side project</strong></a><strong> to discuss and explore the possibilities of </strong><a href="https://booster.cloud/"><strong>Booster</strong></a><strong> to build a top-notch headless e-commerce</strong>. We feel that, more than ever, we should <strong>fight back and give some love to a space that has been lacking from real innovation for too many years</strong>. If it resonates with you, <a href="https://spectrum.chat/boostercloud/commerce/lets-explore-how-to-build-a-modern-event-driven-e-commerce-in-booster~f3755940-a87d-427c-8fc9-2196c34ec38d">come and join us</a>!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=94878c50560a" width="1" height="1" alt=""><hr><p><a href="https://medium.com/the-theam-journey/riding-the-elliot-hype-rollercoaster-and-now-what-94878c50560a">Riding the Elliot hype rollercoaster: and now what?</a> was originally published in <a href="https://medium.com/the-theam-journey">The Agile Monkeys’ Journey</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Asynchronous communication: The key to effective [remote] work.]]></title>
            <link>https://medium.com/the-theam-journey/asynchronous-communication-the-key-to-effective-remote-work-ab53f3c405f4?source=rss-72b6853a68ea------2</link>
            <guid isPermaLink="false">https://medium.com/p/ab53f3c405f4</guid>
            <category><![CDATA[covid19]]></category>
            <category><![CDATA[remote-working]]></category>
            <category><![CDATA[productivity-hacks]]></category>
            <category><![CDATA[asynchronous]]></category>
            <category><![CDATA[productivity]]></category>
            <dc:creator><![CDATA[Javier Toledo]]></dc:creator>
            <pubDate>Tue, 21 Apr 2020 20:51:49 GMT</pubDate>
            <atom:updated>2020-04-21T20:51:49.653Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*fGr1RTGmr_OcGckhRC1shQ.jpeg" /><figcaption>Photo by <a href="https://unsplash.com/@christinhumephoto?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Christin Hume</a> on <a href="https://unsplash.com/s/photos/typing?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Unsplash</a></figcaption></figure><p>Because of the COVID-19 pandemic, a lot of people have been forced to switch to remote work overnight. The natural tendency for someone new to remote work could be to replace the office chatter by constant videoconference calls, but it’s easy that they end up filling your whole workday, leaving you with the feeling that you made no real progress.</p><p>Anyone who has been working remotely for a couple of years knows that to stay productive, <strong>it’s key to adopt an async communication style, reducing meetings and increasing written communication</strong>. But how do you know what to write and when?</p><p>One approach that works for me is:</p><blockquote>Work as if you had no memory and your computer was going to be whipped out over the night</blockquote><ul><li>Every time you work in a task <strong>take notes</strong> of what’s your current status, what you’re thinking and what you plan to do next.</li><li><strong>Publish your notes</strong> in a comment on the corresponding issue in your project tracker. There’s no need to write an essay, just a few words are better than no notes.</li><li>Make sure to <strong>have all your resources permanently online</strong> (Github, Google Drive, Dropbox, shared storage or what makes sense for you) and <strong>include links in your notes to every resource mentioned or needed to continue the work</strong>.</li></ul><blockquote><strong>Side note:</strong> There’s no such thing as oversharing while working remotely, It’s very useful and time-saving to <strong>share links to any resource you mention in any conversation</strong>, even if you know that the other person knows how to locate them. Other team members might not be in the same context, and this practice reduces the friction for them to check your work.</blockquote><ul><li><strong>Use your team chat as a notification center</strong>: Mention that you’ve updated an issue, share links to documents or posts you’ve written or even post the notes there too with links to other resources. Chats tend to receive a lot of attention because they’re too easy to use, but they can easily become a waste of time, so using them to redirect the attention to real work is a nice way to teach better ways of communication by showing.</li><li>And last but not least, <strong>write in a positive way to motivate the rest of the team</strong> with every step of your progress. Sharing a partial success can show progress, and even if you share a problem, you can make it positive with some theories or a list of things you already tried.</li></ul><p>Following these recommendations have many benefits:</p><ul><li><strong>Anyone in your team can see how a task is going</strong>, what are you doing and what you’re planning next, allowing them to intervene or send you tips to make it easier.</li></ul><blockquote>Popular unpopular belief: Daily meetings suck. Twice while working remotely.</blockquote><ul><li>If for any reason you need to reprioritize your work, <strong>it’s easier to restart the task quickly later</strong>.</li><li>If you have to switch to a different project, <strong>you can redirect the task to another person and they’ll have all the context right away</strong>.</li><li>If your computer gets broken, stolen, or lost, <strong>you lose no data nor context</strong>.<em> Extra tip: Be aware of the bad guys, remember to always encrypt your hard drive! ;-)</em></li><li>Taking notes and writing down your current status and plan <strong>helps you to think about what you’re doing </strong>and make better decisions.</li></ul><p>People that are used to have their coworkers on the same room might tend to under-communicate when they switch to remote, when you’re all in the same room it’s easier to keep a shared team-wide context. Losing this shared context can lead to problems ranging from spending a lot of time doing the wrong task (nobody will pass by and ask how you’re doing on their way to the coffee machine), to wasting a lot of time in meetings trying to re-compose the information later. When you don’t write things down the only resource you have is your memory, a non-searchable device with a single I/O bus, so every time that someone needs the information stored in it, you’ll be forced to stop doing whatever you’re doing to communicate it.</p><p>This way of working is extremely inefficient and comes from a clear cognitive bias. It all starts with this feeling that “you’re too busy to write an explanation/comment/note”. You’re trying to avoid 30m of writing when you have all the context in your head to switch to the next task sooner and solve more tasks per unit of time. That could even make you feel like an overachiever, but in reality it’s more like being the one bailing more water out of a sinking boat. The reality is that you’re blocking your team and making them be constantly asking you, becoming a bottleneck and making them play a guess game on if they’ll need a new bit of information 5 minutes after the last clarification call. You could easily find yourself investing hours of Q&amp;A sessions just to avoid those 30m of writing.</p><p>Writing things down saves everyone’s time, starting by yours, and once it’s written and you put it in a shared place, you’ll quickly see your coworkers answering other’s questions with links to your notes. How awesome is that? With a 30m investment you’re not answering to one person, but two, in different timeframes and with zero intervention or extra meetings, and the answers are there for the next one who asks. Time savings are exponential!</p><p>Summarizing: Async communication is awesome, try it and you’ll never go back. Share the resources and context as you generate them and write stuff down for an exponential productivity boost for you and your team. It’s tempting to think that you’re saving time by not writing notes, but the reality is that the alternative is a tsunami of meetings and questions that will quickly exceed the initial time savings. Synchronous communication like meetings and chats are cheap to start but very expensive in the long term because the outcomes are not ready to be reused. And more interestingly, getting used to this approach can make you more productive way after you stop working remotely!</p><p>What else do you consider key to be effective working remotely?</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=ab53f3c405f4" width="1" height="1" alt=""><hr><p><a href="https://medium.com/the-theam-journey/asynchronous-communication-the-key-to-effective-remote-work-ab53f3c405f4">Asynchronous communication: The key to effective [remote] work.</a> was originally published in <a href="https://medium.com/the-theam-journey">The Agile Monkeys’ Journey</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Curtain up!]]></title>
            <link>https://medium.com/the-theam-journey/curtain-up-b31cc4a8bc05?source=rss-72b6853a68ea------2</link>
            <guid isPermaLink="false">https://medium.com/p/b31cc4a8bc05</guid>
            <category><![CDATA[theagilemonkeys]]></category>
            <category><![CDATA[team-building]]></category>
            <category><![CDATA[collaboration]]></category>
            <category><![CDATA[company-culture]]></category>
            <category><![CDATA[bootstrapping]]></category>
            <dc:creator><![CDATA[Javier Toledo]]></dc:creator>
            <pubDate>Tue, 03 Mar 2020 18:06:12 GMT</pubDate>
            <atom:updated>2020-03-03T18:06:12.999Z</atom:updated>
            <content:encoded><![CDATA[<h4>A new post series about <a href="https://www.theagilemonkeys.com/">The Agile Monkeys</a> culture</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*j-lrWnChp9qyCiP2EWtTDA.jpeg" /><figcaption>Photo by <a href="https://unsplash.com/@ryan_hutton_?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Ryan Hutton</a> on <a href="https://unsplash.com/collections/365/night-sky?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Unsplash</a></figcaption></figure><p>From the very beginning, <a href="https://www.theagilemonkeys.com/"><strong>The Agile Monkeys</strong></a><strong> has been a “stealth company.”</strong> We’ve been pretty opaque, writing almost <strong>no public posts</strong>, <strong>not going to many conferences</strong> as either attendees or speakers, and having <strong>a minimalistic website</strong>. This was a decision we made early on, to <strong>spend our time working instead of filling it with what everyone else expects</strong> from a technology company. For many years <strong>we didn’t even have a marketing person</strong>. We invested zero in ads, <strong>growing solely by word of mouth</strong> by <strong>making our clients extremely happy</strong> every time so they would recommend us to our next client. This <strong>has worked splendidly for us</strong>, and we’ve grown from being just a couple of crazy guys to <strong>a company of more than 44 people</strong>, with <strong>consistent annual revenue growth of over 50%</strong> and <strong>year after year profitability</strong>.</p><p><strong>We have never been a startup</strong>, if a startup is something that burns someone else’s money until it either fails or takes off. <strong>We don’t believe in overnight success stories</strong>, either, because <strong>they’re casino stories</strong>. In every casino there are overnight success stories: someone who hit the jackpot one night and became rich. But <strong>99% of the people in casinos lose everything they bet</strong>. Overnight success stories are <strong>what make people keep betting and losing</strong>, because they believe that eventually they can become one of those nice stories. We’ve always run away from that model. <strong>It’s extremely risky</strong>, the probability of success is exceptionally low, and <strong>we have a responsibility to ourselves and our families</strong>, who depend on this company being able to pay our salaries every month.</p><p>Instead, <strong>we chose another direction: bootstrapping</strong>. And now, after nine years, we can say <strong>we’re in a sweet spot</strong>. We <strong>don’t owe anybody anything</strong>, and <strong>we’re still the ones who are defining our future</strong>. In the past few years <strong>we’ve started to open up to the public</strong> a bit: we’ve started to publish some posts, and we’ve gone to a few conferences — not because we’ve changed our minds, but <strong>because they’re a tool for us</strong>. We only go to a conference when we have something to say or when we want to meet someone, and we only post things that we’d write internally anyway. <strong>Our priority is, and always will be, our work, giving the best service possible to our clients, and working hard on our products.</strong></p><blockquote><strong>Our priority is, and always will be, our work, giving the best service possible to our clients, and working hard on our products.</strong></blockquote><p>During all these years <strong>we’ve developed a very special company culture</strong>. We’ve put <strong>our team above everything</strong> else, even money. When there are problems, <strong>we all help</strong>, and when there are successes, <strong>we all feel like we’ve succeeded</strong>. We have a <strong>no-blame culture where anyone is allowed to make mistakes</strong> and the most important thing is how WE fix it, not who caused the problem in the first place. <strong>Everything is built around trust</strong>, to the point that <strong>we don’t care when or where our team members work</strong>: we understand that people’s energy depends on a lot of factors and <strong>all we ask is that they do their best</strong>, because, by definition, nobody can go beyond that. We want <strong>energized people who rest properly and have healthy lives</strong>, and we understand that <strong>we are humans </strong>who are sometimes living through situations that make us less productive. We know that our “best” in those situations will be lower, and we accept that, because <strong>we have a team to help us</strong>. As soon as everyone in the team is doing their best, <strong>the greatest possible outcome will emerge naturally</strong>.</p><p>We’ve had the good luck of finding the most talented, responsible, and hard working colleagues ever, and<strong> they are the reason why our model works</strong>. The<strong> team is the key</strong>, and <strong>we’ve become very skilled at finding more great people and teaching them to work this way</strong>, too. One especially good thing about our culture is that <strong>it’s contagious</strong>, and once you try it <strong>you can’t imagine going back</strong> to the classic 9-to-5 work-in-an-office-space with the false-happiness kindergarten-like perks.</p><p>Most <strong>people don’t believe us</strong> when we talk about how we do things. In fact, some of our colleagues actually tell us after they’ve spent some time with us, “I thought you were lying during the selection process, like all the other companies I worked for did before.”</p><p>The thing is, <strong>we’re realizing that we have something very special</strong> and we feel<strong> it’s time to share it</strong>. First, because we’ve grown to a point where <strong>we can’t keep training every new person individually</strong>, and second, because we think that <strong>it can inspire other people</strong> and also attract prospective clients or future team members to our company.</p><p>So this is the first in a series of articles about <a href="https://www.theagilemonkeys.com/">The Agile Monkeys</a> culture, about how we think, how we do things, and why they’ve worked so well.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=b31cc4a8bc05" width="1" height="1" alt=""><hr><p><a href="https://medium.com/the-theam-journey/curtain-up-b31cc4a8bc05">Curtain up!</a> was originally published in <a href="https://medium.com/the-theam-journey">The Agile Monkeys’ Journey</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Serverless; the next big thing]]></title>
            <link>https://medium.com/the-theam-journey/serverless-for-business-people-dd477959e7c1?source=rss-72b6853a68ea------2</link>
            <guid isPermaLink="false">https://medium.com/p/dd477959e7c1</guid>
            <category><![CDATA[serverless]]></category>
            <category><![CDATA[cloud-computing]]></category>
            <category><![CDATA[faa]]></category>
            <category><![CDATA[engineering-management]]></category>
            <category><![CDATA[lambda-function]]></category>
            <dc:creator><![CDATA[Javier Toledo]]></dc:creator>
            <pubDate>Fri, 14 Sep 2018 18:01:57 GMT</pubDate>
            <atom:updated>2018-10-19T15:34:03.783Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*C1q0ngdNMd0PrWNyz0xI6g.png" /></figure><h4>The technology that will change how we build backend systems forever</h4><blockquote><strong>TL;DR: Serverless</strong> is a new approach to backend software development, where <strong>new systems can be created out of existing third-party services and tools</strong> using lambda functions to glue them together. This <strong>will have a huge impact in how we build software</strong>.</blockquote><p><a href="https://martinfowler.com/articles/serverless.html">Serverless is an evolution of the microservices concept</a>. In a microservices architecture, a big software system is broken into many independent smaller services that work together to bring value. In a Serverless system, instead of having a full project for each microservice where you group code with similar functionality, the backend <strong>functionality is split into a number of independent lambda functions that solve a single use case each</strong>.</p><p>Major cloud providers like <a href="https://aws.amazon.com/lambda/">Amazon Web Services</a>, <a href="https://azure.microsoft.com/en-us/services/functions/">Microsoft Azure</a>, or <a href="https://cloud.google.com/functions/">Google Cloud</a> Platform have <strong>Functions as a Service</strong> (FaaS) products where you can deploy those smaller chunks of code. On those services, <strong>you’re charged exclusively for the execution time of your code</strong>, as opposed to paying for your own reserved server instances, whether or not the servers are idle. A system using lambda functions is, on average,<strong> </strong><a href="https://www.trek10.com/blog/lambda-cost/"><strong>orders of magnitude cheaper than a conventional system</strong></a>.</p><p>In addition to the cost reduction, <strong>lambda functions enable virtually unlimited scalability</strong>. As the cloud provider creates a fresh new function instance for each request, a Serverless system has <strong>no limits regarding the number of users it can serve</strong>. It is as if each user had their own server.</p><p>In addition to that, going <strong>serverless does not require you to drop your existing code</strong> and start from scratch. You can build new features for a legacy app on top of lambda functions and keep the working code running. It is a technology that you can try with <strong>reasonably low risk</strong>.</p><p>For all these reasons, the Serverless approach <a href="https://dashbird.io/blog/companies-using-serverless-in-production/">is being used more and more in the industry</a>, especially by companies with high-traffic needs like <strong>Netflix</strong>, <strong>Nordstrom</strong> or <strong>The Coca-Cola Company</strong>. Concurrently, <a href="https://g.co/trends/tcf5v">the number of developers interested in Serverless keeps growing</a>.</p><p>From the developer perspective, <strong>developing a serverless application is very different than developing a conventional application</strong>. It is easy to start with no extra knowledge, but it is also easy to take the wrong path if you approach your lambda functions as if they were simply smaller microservices. <strong>Deeper study, training, and hiring external expertise should be considered before initiating a Serverless system</strong>. Serverless applications match much better with event-driven approaches, which use different patterns and techniques than the typical Node or Java application.</p><p>We think that <strong>Serverless will play a huge role in the future of backend programming</strong> because it allows companies to <strong>focus on code with business value</strong> and forget operations and glue code concerns. We see this as the difference between building your own chair from scratch, cutting down a tree and doing every single step yourself, and just going to a furniture store and buying a chair. The chair you build yourself might be prettier if you have the skills, but a beautiful hand-crafted chair might not be that valuable to your business if it has nothing to do with chairs. If you only need a chair for your customers to sit on, the time and effort you spent building the chair are just distractions from assisting your customers properly.</p><p>We understand that many companies are still reluctant to put their code in the hands of a cloud provider or let them manage security and scalability for them, but the benefits are so huge and the expertise of the cloud providers, in terms of security and operations, is so vast that <strong>we’re confident that eventually most companies will follow this approach</strong> and stop managing their own servers.</p><p>Do you need help with stepping into Serverless? <a href="/service/mailto:hello@theam.io">Contact us!</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=dd477959e7c1" width="1" height="1" alt=""><hr><p><a href="https://medium.com/the-theam-journey/serverless-for-business-people-dd477959e7c1">Serverless; the next big thing</a> was originally published in <a href="https://medium.com/the-theam-journey">The Agile Monkeys’ Journey</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The future of e-commerce]]></title>
            <link>https://medium.com/the-theam-journey/introducing-the-swing-project-e5ccc73a2c7f?source=rss-72b6853a68ea------2</link>
            <guid isPermaLink="false">https://medium.com/p/e5ccc73a2c7f</guid>
            <category><![CDATA[big-data]]></category>
            <category><![CDATA[microservices]]></category>
            <category><![CDATA[reactive-programming]]></category>
            <category><![CDATA[event-sourcing]]></category>
            <category><![CDATA[ecommerce]]></category>
            <dc:creator><![CDATA[Javier Toledo]]></dc:creator>
            <pubDate>Tue, 04 Oct 2016 15:50:14 GMT</pubDate>
            <atom:updated>2021-05-25T18:53:24.511Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*VWdQ5aJOBEbxp53x6PTpcw.jpeg" /></figure><h4>What would an e-commerce platform look like if we designed it with today’s technology?</h4><p>For the past four years we’ve been developing all kinds of commerce software, including mobile commerce apps, Magento API extensions, in-store applications, funny legacy software integrations, and everything in between.</p><p>It’s interesting that, in spite of <a href="https://www.google.com/trends/explore?q=serverless,microservices">the current trend of serverless and microservices architectures</a> and the introduction of modern and faster languages like Go and Swift, <a href="http://www.thewhir.com/web-hosting-news/magento-woocommerce-lead-ecommerce-platform-market-share-report">more than a half of the e-commerce installed base is built over the classic LAMP stack, featuring monolithic architectures.</a></p><p>At the time of this writing, for instance, Magento is the e-commerce platform with the biggest slice of market share, and there are <a href="https://magento.com/customers/customer-showcase">very big players using it</a>. Thanks to its flexibility and extensibility it can pretty much support any imaginable use case in the e-commerce world. That doesn’t come for free, though: Magento internals are really hard to understand, from both the user and development perspectives, so if the default Magento installation is not enough for you (which is the case for most mid-size+ companies), you’ll invariably need the help of specialized Magento developers and have to rely on third-party extensions and services. In addition, <a href="https://www.keycdn.com/blog/speed-up-magento/">achieving good performance</a> and <a href="http://fbrnc.net/blog/2015/10/super-scaling-magento">scaling</a> or <a href="http://devdocs.magento.com/guides/v2.0/howdoi/migrate/migrate.html">upgrading</a> Magento can be really challenging quests for your team.</p><p>We’ll save a rundown of the strengths and weaknesses of current e-commerce platforms for another post, though. In this article we’d like to share our vision of how a modern commerce system would look based on our experience in this field, discussions we’ve been having with people all around the world, and a great big dose of imagination.</p><h3>Our dream commerce system</h3><blockquote>Most people fail in life not because they aim too high and miss but because they aim too low and hit.</blockquote><blockquote>— Les Brown</blockquote><h4>Microservices architecture</h4><p>Nowadays we could say that commerce applications have a “macroservices” architecture. In order for a retail company to work, it needs to coordinate a number of complex software pieces. As an example, a typical mid-size company can use Magento for its storefront, an ERP backstage to manage suppliers, fulfill orders, and consolidate financial details, and a bunch of scripts and third-party tools integrated along with them.</p><p>We think that breaking those “macroservices” into smaller ones is especially suitable for the commerce domain because it can be easily broken into smaller subdomains that have entities by themselves. For example, a typical e-commerce application would have an online storefront, an order management system, and a fulfillment system to optimize deliveries among other subsystems. These three subsystems can usually work independently, and, when they need to communicate, they can work as a chain using the data produced in one module as the input for the next one.</p><p>We also believe that we need to think about the software systems for a commerce company as a whole — there’s too much clunky proprietary closed software in the market that doesn’t play well with others. We think that the industry needs to switch to a more open approach and we are positive that a microservices approach would be a catalyst for a progressive change in that direction, where businesses would be able to keep using their current systems and progressively replace some of their functionalities as simpler modules are developed.</p><p>Splitting the main submodules of commerce applications into microservices could yield many benefits:</p><ul><li>Smaller, decoupled and domain-focused systems are <strong>easier to maintain</strong> than a huge, monolithic, and generalist codebase.</li><li>The <strong>subsystems can be deployed and scaled independently</strong>, so you could have a lot of machines dedicated to the high-traffic storefront and just a few for your warehouse operations. You can also update one service without the risk of affecting the others.</li><li>The system is <strong>easier to extend and customize</strong>. Well-known APIs between services can be used to create new modules or replace existing ones, without the other modules even noticing. Those APIs make them also <strong>easier to integrate</strong> with other systems.</li><li>The <strong>system becomes independent of a specific technology</strong>, so you can use whichever technology best suits each specific problem.</li></ul><h4>Event-Driven; Reactive!</h4><p><strong>The commerce domain is real-time by nature</strong>: customers are visiting our physical stores and using our web and mobile apps, our employees are processing orders, and providers are delivering new goods, everything at the same time. Being able to react to these events is usually hard for most companies, sometimes impossible, requiring costly integrations and probably a dedicated IT team.</p><p>One of the hardest challenges we’ve found in commerce systems is <strong>keeping everything in sync and consolidated</strong> among all the systems in the company. Every company is unique, with ERPs, online catalogs, or point of sale software, all of them customized for the company’s very specific needs and all using a wide range of third-party services and extensions. It’s not surprising that it’s almost impossible to find solutions that work right out of the box.</p><p>This is not a problem exclusive to the commerce domain. Actually, it’s a more general problem: how to keep coherence in a complex system that is working in a real-time environment. We’ve learned that <a href="https://engineering.linkedin.com/distributed-systems/log-what-every-software-engineer-should-know-about-real-time-datas-unifying">Linkedin had exactly the same problem and they found that moving to an event-driven architecture helped them to decouple their systems and make things easier to manage</a>. Lately, we’ve also found ourselves using solutions like RabbitMQ or Apache Kafka to solve some specific sync problems and the overall feeling from our development team is that they want more of that.</p><p>Front-end development, on the other hand, needs to deal with many real-time events. We’ve been watching event-driven architectures <a href="https://www.google.es/trends/explore?date=all&amp;q=reactivex,reactjs,reactive%20cocoa">like ReactiveX, reactive cocoa, and especially Facebook’s React becoming mainstream</a>, making the industry pay more attention to initiatives like the <a href="http://www.reactivemanifesto.org/">Reactive Manifesto</a>.</p><p>Reactive systems go a step beyond event-driven systems: <a href="https://gist.github.com/staltz/868e7e9bc2a7b8c1f754">they see the event stream as something that can be transformed into new event streams, so <strong>the system is defined in terms of events and transformations instead of data models</strong></a>. We think that this idea can really make the difference when designing commerce systems, and produce simpler and more decoupled subsystems.</p><h4>Schema-less data and heterogeneous databases</h4><p>If there’s something that all commerce companies have in common it’s that they’re all different. While from a high-level view they all look very similar, they have different policies in the way they do things and small but potentially significant differences in the data they manage. For instance, there are companies that have a single photo for each product in their catalog, while others have dozens for each version of the same product; there are companies that allow guest purchases and others that don’t. These small details have a bigger impact at the API/Integration level than you might expect, which is one of the reasons why integrations require that amount of work.</p><p>Most current commerce solutions rely on relational schema-driven databases and, in order to mitigate this problem, they have <a href="http://excellencemagentoblog.com/blog/2011/09/07/magento-eav-database-structure/">configurable database structures that are quite complex,</a> and really hard to work with. In recent years <a href="https://www.quora.com/Why-has-there-been-an-explosion-of-different-data-storage-technologies-in-recent-years">there has been an explosion of NoSQL databases</a>, which <a href="https://dev.mysql.com/doc/refman/5.7/en/document-store.html">have influenced the way the SQL databases have evolved in turn</a>. We believe that switching to a more open schema-less approach could significantly simplify critical parts of commerce software while keeping its flexibility.</p><p>We don’t think of NoSQL databases as a replacement for relational databases, but they’re definitely tools to consider when designing modern systems, because their different approaches may help to solve some specific problems. Using a microservice event-driven architecture makes it easier to use the best tool for each case, as each subsystem can have its own set of databases and use the event store to communicate with other parts of the system.</p><h4><strong>Event Sourcing and CQRS</strong></h4><p>Another of the big challenges in the commerce domain is <strong>traceability</strong>, the ability to know what happened in the software system or the business so decisions can be made or problems fixed. Most current systems focus on modeling the commerce domain, but they don’t usually feature a structured log of operations, so it’s common for some information to get lost in the process.</p><p><a href="http://martinfowler.com/eaaDev/EventSourcing.html">Event sourcing is a technique where a log of committed events is stored and then used to build the system state</a>. This has many implications:</p><ul><li>It’s easy to replay events to rebuild current or previous states. This makes <strong>finding and fixing the origin of problems easier</strong>.</li><li>Subsystems become <strong>more robust</strong>; if for some reason a subsystem loses its state, it can be easily recreated from the event log.</li><li>The event store is an append-only database, making it <strong>easier to scale and manage</strong>.</li><li>By making a copy of the production event store, you can introduce fake events to <strong>predict how the system will behave in certain circumstances.</strong></li></ul><p>In any given event-sourced system it is important to share a standardized approach to generating events, but every system can read the event source and build its own versions of the data in a different way. This is a technique known as <a href="http://martinfowler.com/bliki/CQRS.html">CQRS (Command Query Responsibility Segregation)</a>, which can lead to more independence between subsystems, and data models better suited for the specific problem the submodule needs to solve.</p><h4>Data, data, and data</h4><p>Saving all the events in an event store has an interesting consequence: you have access to all the raw data from your company activities. That means that you can build projections of your data, use data science techniques to study it, and create machine learning or mathematical models. For instance, we’ve found data science and AI to be especially powerful in commerce applications to solve problems like:</p><ul><li>Automating recurrent decisions like stock balancing, when to order new goods from suppliers, or when and where to promote a new product.</li><li>Detecting possible problems before they happen and sending alerts to avert them.</li><li>Improved recommendation systems that take into account a wider range of parameters.</li><li>Optimization of operations in general.</li></ul><p>Thanks to the event stream and the reactive architecture, data-driven modules can be easily introduced without affecting other subsystems.</p><h4>User interface independence</h4><p>Retail companies need user interfaces that match their brands, to enable their customers to identify them and have a great experience. They also need more than a single UI: it’s not enough today to have just a website and mobile apps for the major platforms. We’re seeing more and more companies with <a href="http://www.zdnet.com/article/macys-taps-ibm-watson-to-improve-in-store-shopping-app/">in-store apps</a>, <a href="http://retail-innovation.com/digital-changing-room-and-interactive-mirrors-at-rebbecca-minkoff/">digital dressing rooms</a>, or <a href="http://www.mobilecommercedaily.com/hotels-com-adds-value-to-mobile-bookings-via-smart-watch-integration">smartwatch integrations</a>, and they’re entering new fields like <a href="https://www.amazon.com/Dash-Buttons/b?ie=UTF8&amp;node=10667898011">the internet of things</a> or <a href="http://retail-innovation.com/tommy-hilfigers-in-store-virtual-reality-catwalk-experience/">virtual reality</a>.</p><p>We believe that a modern commerce framework must be designed independently from the UI and focus on the domain logic. UI templates are great for starting companies or low-budget projects, but if you have a great API and use modern tools, it’s usually harder to adapt a configurable UI to your needs than to start a new UI from scratch. In the case of customer-facing applications, it’s usually better to avoid templates if you want to achieve a great user experience. As a modern system, we want to avoid the need for re-engineering the backend in the future, <a href="http://www.bloomberg.com/news/features/2016-05-24/will-this-artificial-reality-machine-really-replace-your-pc">even when augmented reality replaces all our devices.</a></p><h4>Open source!</h4><p>This is the last and most important point. If there’s something we’ve learned working on commerce applications over the last few years, it’s that privative software just makes things harder. Commercial “enterprise” systems tend to make the data captive and are usually non–interoperable: in many cases, the only way to make things work is to pay a fortune to the software supplier to make a custom integration. We believe that we need more open standards and improved interoperability, and we believe the only way to achieve that is to take the open-source path.</p><h3>Update: Let’s make it real!</h3><p>Since I wrote this article in 2016, we’ve been putting these ideas into action, testing all kinds of alternatives on different real-world applications. In 2019, pooling together the knowledge acquired, we unveiled the <a href="https://booster.cloud/">Booster Framework</a>, an ongoing open-source project to easily build event-sourced backend applications on <a href="https://medium.com/the-theam-journey/serverless-for-business-people-dd477959e7c1">Serverless</a> technologies like <a href="https://aws.amazon.com/lambda/">AWS Lambda</a> and <a href="https://aws.amazon.com/dynamodb/">DynamoDB</a>. If you try it or want to collaborate, leave us a message on our <a href="https://discord.gg/bDY8MKx">Discord channel</a>!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=e5ccc73a2c7f" width="1" height="1" alt=""><hr><p><a href="https://medium.com/the-theam-journey/introducing-the-swing-project-e5ccc73a2c7f">The future of e-commerce</a> was originally published in <a href="https://medium.com/the-theam-journey">The Agile Monkeys’ Journey</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The best toolset for a designer]]></title>
            <link>https://medium.com/@javier_toledo/the-best-toolset-for-a-designer-656e1c88df62?source=rss-72b6853a68ea------2</link>
            <guid isPermaLink="false">https://medium.com/p/656e1c88df62</guid>
            <dc:creator><![CDATA[Javier Toledo]]></dc:creator>
            <pubDate>Sun, 03 Nov 2013 23:52:01 GMT</pubDate>
            <atom:updated>2013-11-04T00:43:52.792Z</atom:updated>
            <content:encoded><![CDATA[<h4>An unexpected way to discover it :-)</h4><p>A few days ago, one of our programmers sent this image through internal mail, “The perfect designer workstation”:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/500/1*WSt_FMZ5DiBi6OkykwFOgQ.jpeg" /></figure><p>It was fun, everyone smiled and continued working, but some hours later, Alejandro, our chief designer and one of the best designers I’ve ever met responded with this short but glorious quote:</p><blockquote>No, the best toolset for a designer is this one:</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/450/1*hg0T1_q521WtCay7Jf-I3w.jpeg" /></figure><p>At this moment I realized that I haven’t ever seen Alejandro coming to a meeting with a laptop, just an iPad to search on the web and this tiny notepad where he take notes and draw little sketches, and which he carries with him wherever he goes. He is one of these designers that build and test ideas with pencil and paper before sitting at his computer, and I think that this is one of the details that make him so special.</p><p>Maybe sometimes we focus too much on the tools we use, losing the perspective on what we really have to do. Maybe the best toolset for a designer is not a futuristic science-fiction impressive workstation that does it all for us with a keystroke, but the freedom to play, test and imagine things beyond the boundaries of tools.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=656e1c88df62" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>