<?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 Salih KILIÇ on Medium]]></title>
        <description><![CDATA[Stories by Salih KILIÇ on Medium]]></description>
        <link>https://medium.com/@salihklc91?source=rss-7142079a79f------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*fVmJpmY5q6VGAQjDzQtRNg.jpeg</url>
            <title>Stories by Salih KILIÇ on Medium</title>
            <link>https://medium.com/@salihklc91?source=rss-7142079a79f------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Tue, 17 Nov 2026 22:54:12 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@salihklc91/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[Sustainable Efficiency]]></title>
            <link>https://medium.com/@salihklc91/sustainable-efficiency-fba77f4c96f3?source=rss-7142079a79f------2</link>
            <guid isPermaLink="false">https://medium.com/p/fba77f4c96f3</guid>
            <category><![CDATA[future-of-technology]]></category>
            <category><![CDATA[software-engineering]]></category>
            <category><![CDATA[team-efficiency]]></category>
            <dc:creator><![CDATA[Salih KILIÇ]]></dc:creator>
            <pubDate>Tue, 11 Mar 2025 08:24:12 GMT</pubDate>
            <atom:updated>2025-03-11T08:24:12.887Z</atom:updated>
            <content:encoded><![CDATA[<p><strong>Software Development Teams and the Future of Technology</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ICVMDeM8DtCQ99tedyjN-g.png" /></figure><p>Hello there 👋🏼</p><p>I’ve finally completed the article that combines my impressions from the GOTO 2024 conference with my own observations — something I’ve been thinking about for a long time, but could never find the time to focus on. In this article, I will discuss how we try to be achieve efficiency in software teams and applications, looking at it not just from a performance standpoint but from multiple angles. I hope you enjoy reading!</p><p>The tech world continues to move forward at full speed. Each year, new concepts, technological breakthroughs, and improvement methods emerge. On the one hand, these make our way of working easier; on the other hand, they spur us to constantly want “more.” As software developers, each application we build aims for higher performance, faster response times, and far lower energy consumption. At this point, we see that the concept of efficiency expands well beyond just coding, touching on areas such as team structures, project management, hardware constraints, and even sustainability. (Some might argue that this is merely a “byproduct of capitalism,” but we’ll leave that debate for another article.)</p><p>In this piece, I’ll try to take a broader perspective on our software development processes, guided by the insights I gained from <strong>GOTO Copenhagen 2024</strong>. In particular, I want to stress how enhancing efficiency in software teams and the applications we build has a positive impact on both the planet and an organization’s performance. Also, given that <a href="https://en.wikipedia.org/wiki/Moore%27s_law"><strong>Moore’s Law</strong></a> is slowing down, I intend to discuss the advantages of developing more efficient software rather than waiting on more powerful hardware.</p><h3>1. Efficiency and Responsibility: Our Engineering Duty</h3><p>On the first day of the conference, Barbara Oakley’s talk titled <strong>“</strong><a href="https://gotocph.com/2024/sessions/3168/using-generative-ai-to-strengthen-and-speed-learning"><strong>USING GENERATIVE AI TO STRENGTHEN AND SPEED LEARNING</strong></a><strong>”</strong> focused on how artificial intelligence — especially generative AI — can simplify our learning processes. One of the most striking takeaways from Oakley’s presentation was the similarity between how the human brain transitions from a declarative stage of learning to unconscious application (automation) and how <strong>deep learning models</strong> learn to establish connections based on “examples.” This once again reinforced the idea that, as developers, we must adopt a “<strong>learning how to learn</strong>” and “<strong>continuous improvement</strong>” mindset. By delegating the tasks that waste our energy to <strong>generative AI</strong> tools, we can free up our time for more creative work.</p><p>On the other hand, when we hear “<strong>efficiency</strong>,” we usually think only of performance: speeding up code, reducing CPU usage, optimizing memory consumption, etc. Yet the lesson I took from Oakley’s talk is that <strong>efficiency and learning must go hand in hand</strong>. <strong>Efficiency</strong> is directly linked not just to how applications run, but also to how <strong>teams</strong> operate overall. For example, onboarding new members into established, complex systems can be quite challenging. Here, tools like generative AI can improve access to information, boosting efficiency for both teams and individuals. However, if we let these tools push us toward “<strong>laziness</strong>,” we may see a long-term decline in both the team’s expertise and its problem-solving abilities.</p><p>Consequently, engineering isn’t just about “using” new technologies — <strong>it’s about understanding the underlying principles</strong> and applying them in the <strong>most efficient</strong> way possible. Individual responsibilities matter just as much as teamwork responsibilities: beyond simply optimizing the performance of the code we write, we also need to make the <strong>energy we consume</strong> — both physically and mentally — more efficient. This approach is key to becoming a <strong>sustainable team</strong> over the long term.</p><figure><img alt="image is generated by AI with too many errors but in the end main idea is okay, its not valuable if the product does not works." src="https://cdn-images-1.medium.com/max/1024/1*eQSE01gsCOGmVpWejH-Zvg.png" /><figcaption>The image is generated by AI with too many errors/corruption like you see but in the end main idea is okay :) It’s not valuable if the product does not works.</figcaption></figure><h3>2. Team Efficiency: Technical Coaching and Team Topologies, Project or Product?</h3><p>Another major topic highlighted at the conference was <strong>technical coaching</strong> and the ability of teams to self-organize. In her session titled <strong>“</strong><a href="https://gotocph.com/2024/sessions/3116/i-dont-need-another-scrum-master-get-me-a-technical-coach"><strong>I don’t need a scrum master, get me a technical coach</strong></a><strong>,”</strong> Emily Bache explained that software teams’ efficiency often isn’t just a matter of project management or having a “scrum master.” The real problem lies in the need for qualified people who can address code quality and technical challenges.</p><p>As Bache put it, <strong>“You can’t solve problems by looking at Jira; you can only visualize them.”</strong> So, what is a technical coach? Technical coaches not only equip the team with software practice skills but also help new developers adapt quickly to the system. In large teams or where complex architectures need to be managed, <strong>having good developers alone is often not enough</strong>; to maintain ongoing efficiency, you need experienced individuals with strong technical and social skills to guide the team.</p><p>Similarly, Manuel Pais’s talk <strong>“</strong><a href="https://gotocph.com/2024/sessions/3322/from-strategy-to-practice-insights-on-how-team-topologies-drive-organizational-success"><strong>From Strategy to Practice</strong></a><strong>”</strong> introduced the concept of <strong>Team Topologies</strong>, which allows teams to reduce the chaos around them and deliver value more quickly. Structuring teams around products or streams helps minimize dependencies, thereby increasing development speed. Pais notes that <strong>44%</strong> of developers working in teams with high organizational dependencies experience burnout, while well-designed organizations — where dependencies are reduced — see <strong>over 60% improvement</strong> in performance.</p><p>Pais describes four different team types — <strong>stream-aligned team, enabling team, complicated subsystem team, platform team</strong> — and explains how their interactions can help control organizational complexity.</p><blockquote>Here’s a great article in a similar vein: <a href="https://medium.com/bbc-product-technology/refactor-organisation-80e4e171d922"><strong>Refactor → Organisation</strong>.</a></blockquote><p>In addition, the “<a href="https://gotocph.com/2024/sessions/3147/don-t-feed-the-pigeons-some-principles-from-real-world-internal-developer-platform-engineering"><strong>Don’t Feed the Pigeons</strong></a>” session by <strong>Russ Miles</strong> challenged our conventional view of a product and the role of a technical product owner. Miles urged us to “<strong>change your perspective</strong>” — to see every project not <strong>just as a task, but as a product</strong> that should be treated with the same care as any market offering. He emphasized principles like “<strong>make the right thing simple and easy</strong>” and “<strong>least surprise</strong>”, reminding us that a product’s interface and code must communicate clearly with its users. Success lies not in rigid structures or titles, but in behavior, cultural alignment, and the willingness to listen to the users. By actively using our own products — essentially “eating our own dog food” — we can directly experience what works and what needs improvement. This experimental, person-centered approach to product stewardship not only meets the creative autonomy of our teams but also ensures that our work continually resonates with its intended audience. <strong>In short, by treating every project as a product and engaging with it from the user’s perspective, we can drive sustainable team efficiency and truly make an impact.</strong></p><p><strong>Summing up</strong>, an efficient team is one that:</p><ul><li><strong>Has minimized dependencies</strong></li><li><strong>Has easy access to technical coaches or similar roles</strong></li><li><strong>Uses appropriate tools that facilitate rapid adaptation and learning</strong></li><li><strong>Can handle its own tasks (coding, testing, DevOps, etc.) end-to-end</strong></li><li><strong>Be a user of your own product (not project)</strong></li></ul><p>Such an organization, while busy shipping products, also makes the most of the resources at its disposal. In our own teams, we should focus on addressing issues with these points first.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Ftenor.com%2Fembed%2F21178977&amp;display_name=Tenor&amp;url=https%3A%2F%2Ftenor.com%2Fview%2Fteamwork-gif-21178977&amp;image=https%3A%2F%2Fmedia.tenor.com%2Fy9C-bqGhlloAAAAN%2Fteamwork.png&amp;type=text%2Fhtml&amp;schema=tenor" width="600" height="400" frameborder="0" scrolling="no"><a href="https://medium.com/media/9956ad6a88d68f079fc7b6b82de5a60a/href">https://medium.com/media/9956ad6a88d68f079fc7b6b82de5a60a/href</a></iframe><h3>3. Moore’s Law and the Hardware “Magic” We Can No Longer Rely On</h3><p>For a long time, there’s been a perception that <strong>hardware is a “savior.”</strong> We tackled many problems by moving our applications to more powerful servers or adding more memory/CPU; when that wasn’t enough, we split them in two, or three, and scaled horizontally. As organizational complexity mounted, we broke our services down even further and called them “microservices.” Then we deployed those microservices with container orchestration tools like Kubernetes, turning even simple CRUD apps into massive architectures capable of responding in under 50ms.</p><blockquote>Yes, sometimes we <em>did</em> create piles of garbage — partly joking, partly serious. While some services truly needed this level of scale, many did not.</blockquote><p>Sophie Wilson’s talk, <strong>“The Future of Microprocessors,”</strong> showed that the assumption <em>“hardware will always get faster anyway”</em> no longer holds as true. We are now approaching physical limits for microprocessors; manufacturing at the nanometer scale is becoming increasingly difficult, and costs are skyrocketing.</p><p>The <strong>know-how that companies, which have been in the sector for years, possess has raised the “initial barrier to entry”</strong> in the processor industry. A few years ago, there were more than <strong>20 companies</strong> that manufactured processors, but now <strong>only a handful</strong> can operate at a modern scale. On one side, Intel and Qualcomm can produce at <strong>4 nm</strong>, while Samsung and TSMC can produce at <strong>3 nm</strong>. Even among these giants, the gap continues to widen. Working at these extremely small scales also introduces additional physical challenges, such as <strong>overheating and cooling requirements</strong>, making production costs soar beyond reason.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Ftenor.com%2Fembed%2F25343056&amp;display_name=Tenor&amp;url=https%3A%2F%2Ftenor.com%2Fview%2Fcooking-cpu-hot-meat-yummy-gif-25343056&amp;image=https%3A%2F%2Fmedia.tenor.com%2F8wsVD890uVkAAAAN%2Fcooking-cpu.png&amp;type=text%2Fhtml&amp;schema=tenor" width="600" height="400" frameborder="0" scrolling="no"><a href="https://medium.com/media/c7eef418b6a3541feeceb71052bafc60/href">https://medium.com/media/c7eef418b6a3541feeceb71052bafc60/href</a></iframe><p>Perhaps the classic mindset inspired by <strong>“Moore’s Law” — that hardware will always get faster, so let’s develop however we like, even if inefficient” — led to the accumulation of technical debt</strong>, especially during the pandemic, when quick fixes became a frequent shortcut. However, <a href="https://en.wikipedia.org/wiki/Amdahl%27s_law"><strong>Gene Amdahl’s Law</strong></a> tells us that <strong>the benefit of adding more cores only increases in proportion to how parallelizable the software really is.</strong> In short, the era of “free speedups from hardware” has come to an end. <em>(This could be a good place to include Moore’s chart.)</em></p><p>Wilson emphasized that <strong>the antidote</strong> to being unable to produce faster processors — or to producing them at prohibitively high cost — <strong>lies in software efficiency.</strong> By developing software that can accomplish the same tasks with <strong>less memory, less processing power, and less energy</strong>, we can <strong>turn the slowdown of Moore’s Law into an advantage.</strong> These constraints can also encourage software teams to focus more on innovation.</p><h3>4. AI Again, Microservices, and Other Tools: Levers for Efficiency</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/735/1*H7nYt5D-d1ZXj8etpIRoew.jpeg" /></figure><h3>Artificial Intelligence (Generative AI)</h3><p>We’ve touched on the main ideas above, but let’s reinforce them here. The notion that AI will take over our daily work came up frequently at the conference. However, the general consensus was that <strong>“AI is not going to replace our jobs; rather, it will help us become more productive.”</strong> AI-assisted code completion, refactoring suggestions, or <strong>test scenario </strong>generation tools (<strong>Not unit test generation</strong>) give developers time back and let them focus on more complex challenges.</p><blockquote>A great example related to the question “Will we be out of jobs when AI arrives?” is <strong>Jevons’ Paradox</strong>. In 1865, William Stanley Jevons argued that more efficient steam engines would not reduce coal consumption, but instead increase it — because as fossil fuels became cheaper, demand would rise, and more engines would be built. The coal usage scenario of the 19th century applies to AI today: New tools increase efficiency but also boost demand. When steam engines became more efficient, the only job that disappeared was for elevator operators; likewise, when transportation cards (Akbil:) appeared, only the staff who collected bus fares lost their roles. But new jobs always emerged to replace the old ones. For instance, in the early 20th century, when cars became widespread, they dramatically reduced the use of horse-drawn carriages. This did not completely eliminate the “driver” profession — only changed its vehicles and required skill sets. <strong><em>For now</em></strong><em>,</em> there’s no need to panic.</blockquote><p>Our mission, then, is to <strong>guide AI tools properly</strong>. If we fail at this, instead of “efficiency,” we might produce “<strong>unnecessary code blocks</strong>” Some speakers noted that code automatically generated by AI tools fares poorly in refactoring processes. This underscores the importance of developers reviewing AI-generated code and staying in a continual learning loop.</p><h3>Microservices</h3><p><strong>James Lewis</strong> and others discussed how the “microservices” approach has evolved over the past decade. Large monolithic systems can become unwieldy, both in terms of the size of teams and the complexity of the code. Thanks to features like<strong> modularity, scalability, and independent deployment</strong>, microservices have improved efficiency. Each service focuses on its own bounded context, can be updated quickly, and as the number of teams grows, the system doesn’t “slow down” but rather is divided into more parts that remain <strong>manageable</strong>.</p><p>Afterward, it was pointed out that in the end, distributed systems offload their complexity onto the <strong>infrastructure</strong>. A frequently asked question throughout the conference was, <strong>“Have we just shifted all complexity from the applications to the infrastructure?”</strong> While the adoption of DevOps culture and self-service platforms has allowed us to feel the efficiency benefits of microservices, new challenges have also arisen on the management side. The past few years have seen the emergence of hundreds of PaaS solutions, container management tools, and auxiliary applications to address these needs. <em>(Examples here could include “Istio,” “K8s,” etc.)</em></p><h3>Domain-Driven Design (it’s just a tool)</h3><p>DDD provides a good approach to modeling complex business problems. When combined with Simon Wardley’s “<a href="https://www.wardleymaps.com/">Wardley Maps,</a>” you can figure out which components are truly <strong>“core domains”</strong> for the business and which are basically <strong>“commodities.”</strong> This allows you to develop high-value areas in-house and outsource or share general/supporting parts with platform teams. Ultimately, this leads to cleaner, more focused, and more efficient software.</p><p>Especially during aggressive growth periods, splitting up projects and expanding teams that handle a particular domain might incur extra costs, but it’s acceptable. However, in more static phases, problems like high maintenance costs, resource consumption, and a “microservice explosion” can surface. Tools like DDD, Team Topologies, and Wardley Maps help identify which parts are better suited for <strong>“outsourcing” </strong>or even<strong> “re-monolithization”. </strong>After all, engineering is the art of creating effective solutions with limited resources.</p><blockquote>The fine line between “projects” and “products” also becomes clear here: one is an area where we test our technological curiosities, while the other is where we deliver sustainable, profitable solutions.</blockquote><h3>5. The Sustainability Dimension: Our Planet and the Reality of “E-Waste”</h3><p>In her talk, <strong>“</strong><a href="https://gotocph.com/2024/sessions/3327/the-efficiency-paradox-and-how-to-save-yourself-and-the-world"><strong>The Efficiency Paradox and How to Save Yourself and the World</strong></a><strong>,”</strong> Holly Cummins emphasized another side of efficiency: <strong>our responsibility to the planet</strong>. The software world sees an astonishing amount of wasted resources — projects that are built and then forgotten, “<strong>zombie servers</strong>” left running without a purpose, and vast quantities of unused data. According to some statistics, about <strong>25%</strong> of servers worldwide are “zombies”, <strong>only 12–18%</strong> of active servers’ capacity is actually utilized, and <strong>68%</strong> of recorded data is never accessed. All of this translates into enormous energy waste and carbon emissions.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Ftenor.com%2Fembed%2F8220789&amp;display_name=Tenor&amp;url=https%3A%2F%2Ftenor.com%2Fview%2Foffice-server-power-lady-unplug-gif-8220789&amp;image=https%3A%2F%2Fmedia.tenor.com%2FbGCuW8uql2kAAAAN%2Foffice-server.png&amp;type=text%2Fhtml&amp;schema=tenor" width="600" height="400" frameborder="0" scrolling="no"><a href="https://medium.com/media/63e943ab99710bde6e57b4e15fb0d7b1/href">https://medium.com/media/63e943ab99710bde6e57b4e15fb0d7b1/href</a></iframe><p>Even simple measures can yield major gains: for example, shutting down applications that have no users at night, using a “<strong>light switch ops</strong>” approach to spin up systems only when needed, or refraining from storing data that’s not actually required. At a more granular level, writing everyday code in a way that avoids unnecessary data storage, applying best practices to reduce memory and CPU usage, or choosing more efficient languages/frameworks (e.g., Quarkus) can significantly benefit both company budgets and the planet.</p><p>A point that conference speakers emphasized over and over is that <strong>efficiency is not just about corporate profitability</strong>; it’s also a key to <strong>building a sustainable future</strong>.</p><h3>6. Conclusion: Efficiency, Innovation, and Responsibility</h3><p>There were many ideas at GOTO 2024 that were worth sharing. Putting them all together, we see that:</p><ol><li><strong>Efficiency isn’t only about performance.</strong> It also relates to how teams organize themselves, how quickly they can resolve issues, and how easily new members adapt.</li><li><strong>We’ve reached the end of relying on Moore’s Law.</strong> In a future where hardware won’t keep getting faster indefinitely, developing more efficient software is a critical success factor.</li><li><strong>Team organization</strong> — whether via Team Topologies, technical coaching, or domain-driven design — <strong>affects not just the technical efficiency of the software but also team efficiency.</strong> Well-structured teams reduce dependencies, boosting both speed and quality.</li><li><strong>AI tools, microservices, DDD,</strong> and similar approaches can be powerful levers to increase efficiency <strong>if used responsibly and with control</strong>.</li><li><strong>Sustainability must be considered</strong>, including energy consumption, resource usage, and carbon footprint. The benefits of efficiency are <strong>not just economic, but also ecological</strong>.</li></ol><p>In the end, as engineers, our task is to use resources <strong>in the best possible way</strong> and choose <strong>the right methods</strong>. Writing efficient code, selecting the appropriate architecture, organizing teams in a way that yields tangible benefits, and continually learning how to adopt new technologies responsibly — all of these are critical, as they affect both our organizations’ competitiveness and the future of our planet.</p><p>Let’s not forget that every application we develop and every line of code we add represents <strong>a great opportunity for building a cleaner and more sustainable future</strong>.</p><h3>References &amp; Further Reading</h3><ul><li><strong>Barbara Oakley</strong> — <a href="https://gotocph.com/2024/speakers/3277/barbara-oakley"><em>Using Generative AI to Strengthen and Speed Learning</em></a></li><li><strong>Emily Bache</strong> — <a href="https://gotocph.com/2024/sessions/3116/i-dont-need-another-scrum-master-get-me-a-technical-coach"><em>The Samman Method, I Don’t Need Another Scrum Master, Get Me a Technical Coach!</em></a></li><li><strong>Manuel Pais &amp; Matthew Skelton</strong> — <a href="https://teamtopologies.com/"><em>Team Topologies</em></a><em>, </em><a href="https://gotocph.com/2024/sessions/3322/from-strategy-to-practice-insights-on-how-team-topologies-drive-organizational-success"><em>From Strategy to Practice: Insights on How Team Topologies Drive Organizational Success</em></a></li><li><strong>Simon Wardley</strong> — <a href="https://gotocph.com/2024/sessions/3128/x-marks-the-spot-navigating-possible-futures"><em>X Marks the Spot: Navigating Possible Futures</em></a></li><li><strong>Russ Miles</strong> —<a href="https://gotocph.com/2024/sessions/3147/don-t-feed-the-pigeons-some-principles-from-real-world-internal-developer-platform-engineering"> <em>Don’t Feed the Pigeons: Some Principles From Real World Internal Developer Platform Engineering</em></a></li><li><strong>Sophie Wilson</strong> — <a href="https://gotocph.com/2024/sessions/3327/the-efficiency-paradox-and-how-to-save-yourself-and-the-world"><em>The Future of Microprocessors</em></a></li><li><strong>Holly Cummins</strong> — <a href="https://gotocph.com/2024/sessions/3327/the-efficiency-paradox-and-how-to-save-yourself-and-the-world"><em>The Efficiency Paradox</em></a></li><li><strong>Kent Beck</strong> — <a href="https://gotocph.com/2024/sessions/3410/tidy-first-a-daily-exercise-in-empirical-design"><em>Tidy First? A Daily Exercise in Empirical Design</em></a></li><li><strong>James Lewis</strong> — <a href="https://gotocph.com/2024/sessions/3331/the-magic-of-small-things-10-years-of-microservices"><em>The Magic of Small Things — 10 Years of Microservices</em></a></li><li><strong>Wikipedia — </strong><a href="https://en.wikipedia.org/wiki/Moore%27s_law">Moore’s Law</a>, <a href="https://en.wikipedia.org/wiki/Amdahl%27s_law"><em>Gene Amdahl’s Law</em></a></li></ul><p>I hope this article provides the developer community and the broader tech world with a new perspective on <strong>efficiency, organizational success, and sustainability</strong>.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=fba77f4c96f3" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>