https://thisisfatih.com Thoughts on software, systems, and good practice. en Sun, 08 Nov 2026 11:57:04 GMT https://thisisfatih.com/posts/five-years-as-a-tech-executive https://thisisfatih.com/posts/five-years-as-a-tech-executive Mon, 20 Apr 2026 00:00:00 GMT Five lessons from five years of hiring, firing, building, and occasionally breaking things. https://thisisfatih.com/posts/good-practices-that-actually-matter https://thisisfatih.com/posts/good-practices-that-actually-matter Tue, 07 Apr 2026 00:00:00 GMT The ones that quietly determine whether a codebase is healthy at year three. https://thisisfatih.com/posts/the-non-linear-path-msc-thesis-to-cto https://thisisfatih.com/posts/the-non-linear-path-msc-thesis-to-cto Tue, 10 Mar 2026 00:00:00 GMT Not a career advice post. A map of how the dots actually connected. https://thisisfatih.com/posts/how-we-use-ai-to-write-and-review-code https://thisisfatih.com/posts/how-we-use-ai-to-write-and-review-code Tue, 17 Feb 2026 00:00:00 GMT The workflow. The guardrails. The moments it goes wrong. https://thisisfatih.com/posts/what-im-watching-in-2026 https://thisisfatih.com/posts/what-im-watching-in-2026 Tue, 20 Jan 2026 00:00:00 GMT The things that look small now but probably aren't. https://thisisfatih.com/posts/ten-years-of-building-with-ai-2015-to-2025 https://thisisfatih.com/posts/ten-years-of-building-with-ai-2015-to-2025 Mon, 08 Dec 2025 00:00:00 GMT What the decade actually looked like from inside it. Not the hype arc. https://thisisfatih.com/posts/what-scales-without-headcount https://thisisfatih.com/posts/what-scales-without-headcount Mon, 06 Oct 2025 00:00:00 GMT The architecture decisions, culture choices, and AI tooling that let a small team carry a large product. https://thisisfatih.com/posts/the-agentic-engineering-loop https://thisisfatih.com/posts/the-agentic-engineering-loop Mon, 22 Sep 2025 00:00:00 GMT From spec to code review to deployment. Where agents live. What they do. What still needs humans. https://thisisfatih.com/posts/why-i-dont-have-a-separate-ai-team https://thisisfatih.com/posts/why-i-dont-have-a-separate-ai-team Mon, 11 Aug 2025 00:00:00 GMT AI is not a feature. It's not a team. It's how the whole team works. https://thisisfatih.com/posts/99-point-9-percent-uptime-with-a-six-person-team https://thisisfatih.com/posts/99-point-9-percent-uptime-with-a-six-person-team Mon, 30 Jun 2025 00:00:00 GMT Operational discipline > headcount. The specific practices. https://thisisfatih.com/posts/model-evaluation-in-production https://thisisfatih.com/posts/model-evaluation-in-production Mon, 19 May 2025 00:00:00 GMT The metrics that matter. The ones that feel good but aren't. https://thisisfatih.com/posts/agentic-workflows-what-works-and-what-blows-up https://thisisfatih.com/posts/agentic-workflows-what-works-and-what-blows-up Mon, 28 Apr 2025 00:00:00 GMT Multi-agent in a real product. The failure modes. The wins. The things you only learn by running it. https://thisisfatih.com/posts/building-a-small-team-that-punches-above-its-weight https://thisisfatih.com/posts/building-a-small-team-that-punches-above-its-weight Mon, 17 Mar 2025 00:00:00 GMT Hiring philosophy, culture decisions, and the AI layer that multiplies output. https://thisisfatih.com/posts/the-claude-api-in-production https://thisisfatih.com/posts/the-claude-api-in-production Mon, 10 Feb 2025 00:00:00 GMT Why I use it, what it's better at, and what I still reach for other tools for. https://thisisfatih.com/posts/hiring-for-startups-takeaways-from-brian-chesky https://thisisfatih.com/posts/hiring-for-startups-takeaways-from-brian-chesky Mon, 20 Jan 2025 00:00:00 GMT Chesky's hiring framework is direct, no-nonsense, and more relevant to startup founders than anything I've read in an HR playbook. https://thisisfatih.com/posts/kritis-compliance-from-inside-an-engineering-team https://thisisfatih.com/posts/kritis-compliance-from-inside-an-engineering-team Mon, 16 Dec 2024 00:00:00 GMT Compliance requirements land on engineering teams whether they asked for them or not. Here's what that actually looks like. https://thisisfatih.com/posts/from-head-of-engineering-to-cto https://thisisfatih.com/posts/from-head-of-engineering-to-cto Mon, 25 Nov 2024 00:00:00 GMT What changed. What being CTO actually means at this stage of the company. https://thisisfatih.com/posts/cutting-costs-30-percent-while-growing-usage https://thisisfatih.com/posts/cutting-costs-30-percent-while-growing-usage Mon, 14 Oct 2024 00:00:00 GMT Specific decisions. Not principles. https://thisisfatih.com/posts/gcp-cloud-run-fastapi-the-stack https://thisisfatih.com/posts/gcp-cloud-run-fastapi-the-stack Mon, 30 Sep 2024 00:00:00 GMT Why this combination. What it costs. What breaks. What I'd change. https://thisisfatih.com/posts/the-okr-system-that-worked https://thisisfatih.com/posts/the-okr-system-that-worked Mon, 19 Aug 2024 00:00:00 GMT I've seen OKRs done badly. Here's the version that worked under real pressure. https://thisisfatih.com/posts/multi-agent-systems-what-actually-works https://thisisfatih.com/posts/multi-agent-systems-what-actually-works Mon, 08 Jul 2024 00:00:00 GMT Not the research version. The version that has to not fail inside a real product. https://thisisfatih.com/posts/growing-from-2-to-30-installations https://thisisfatih.com/posts/growing-from-2-to-30-installations Mon, 24 Jun 2024 00:00:00 GMT The engineering decisions that scaled. The ones that didn't. https://thisisfatih.com/posts/yolo-on-physical-infrastructure https://thisisfatih.com/posts/yolo-on-physical-infrastructure Mon, 13 May 2024 00:00:00 GMT Waste monitoring in the field. Cameras, weather, variable light. The real world is messier than the dataset. https://thisisfatih.com/posts/what-i-found-when-i-joined-wasteer https://thisisfatih.com/posts/what-i-found-when-i-joined-wasteer Mon, 22 Apr 2024 00:00:00 GMT Product directing tech. No ownership. Three-hour meetings. The first 30 days. https://thisisfatih.com/posts/three-years-of-cv-in-production https://thisisfatih.com/posts/three-years-of-cv-in-production Mon, 04 Mar 2024 00:00:00 GMT Not the architecture. The judgment. The things you can only learn by being on call when it breaks. https://thisisfatih.com/posts/technical-due-diligence-from-both-sides https://thisisfatih.com/posts/technical-due-diligence-from-both-sides Mon, 15 Jan 2024 00:00:00 GMT I've been diligenced and I've diligenced others. What the process actually surfaces. https://thisisfatih.com/posts/2023-the-year-the-ceiling-came-off https://thisisfatih.com/posts/2023-the-year-the-ceiling-came-off Sun, 31 Dec 2023 00:00:00 GMT A practitioner's year in review. Not a predictions post. What actually changed, what's oversold, and what I'm watching. What actually changed The interface layer. ChatGPT was November 2022. The consequence landed in 2023. Every enterprise client we work with has had a board level conversation about AI this year. The language model capability was there… https://thisisfatih.com/posts/ai-assisted-sprint-planning https://thisisfatih.com/posts/ai-assisted-sprint-planning Mon, 20 Nov 2023 00:00:00 GMT Not a concept post. Here's exactly what we do, what the AI produces, and what we do with it. The input At the start of sprint planning, we feed the model three things: the product backlog items prioritized for this sprint, the sprint goal drafted by the product owner, and the last three sprint retrospective… https://thisisfatih.com/posts/multimodal-models-and-computer-vision-products https://thisisfatih.com/posts/multimodal-models-and-computer-vision-products Mon, 30 Oct 2023 00:00:00 GMT GPT-4V launched. After three years building computer vision products, here's what I think it actually means. https://thisisfatih.com/posts/prompt-engineering-is-engineering https://thisisfatih.com/posts/prompt-engineering-is-engineering Mon, 25 Sep 2023 00:00:00 GMT The dismissal is wrong. Here's why prompt engineering is a real engineering discipline. https://thisisfatih.com/posts/benchmarks-vs-production https://thisisfatih.com/posts/benchmarks-vs-production Mon, 14 Aug 2023 00:00:00 GMT MMLU scores don't tell you what happens when your prompt is ambiguous at 11pm and the model confidently gives you the wrong answer. https://thisisfatih.com/posts/why-i-switched-from-gpt4-to-claude https://thisisfatih.com/posts/why-i-switched-from-gpt4-to-claude Mon, 03 Jul 2023 00:00:00 GMT Not a hot take. A practical decision made after running both in production for the same tasks. https://thisisfatih.com/posts/the-compliance-problem-with-llms https://thisisfatih.com/posts/the-compliance-problem-with-llms Mon, 22 May 2023 00:00:00 GMT Our enterprise clients can't send data to OpenAI. For several of them, it's a legal constraint. Here's how we solved it. https://thisisfatih.com/posts/ai-assisted-code-review-the-system-we-built https://thisisfatih.com/posts/ai-assisted-code-review-the-system-we-built Mon, 10 Apr 2023 00:00:00 GMT Six months of AI-assisted code review in production. The architecture, the results, and the things it still misses. https://thisisfatih.com/posts/llms-as-infrastructure-not-features https://thisisfatih.com/posts/llms-as-infrastructure-not-features Mon, 27 Feb 2023 00:00:00 GMT Bolting an LLM onto a product as a feature and announcing 'now with AI' is not a strategy. https://thisisfatih.com/posts/six-weeks-of-gpt-4-in-our-workflow https://thisisfatih.com/posts/six-weeks-of-gpt-4-in-our-workflow Mon, 16 Jan 2023 00:00:00 GMT Early API access to GPT-4. Six weeks in, here's the honest account. https://thisisfatih.com/posts/chatgpt-launched-here-is-what-i-actually-think https://thisisfatih.com/posts/chatgpt-launched-here-is-what-i-actually-think Wed, 30 Nov 2022 00:00:00 GMT Having run GPT-3 in production since mid-2020, here's an honest read on what ChatGPT actually is. https://thisisfatih.com/posts/speed-as-a-product-feature https://thisisfatih.com/posts/speed-as-a-product-feature Mon, 05 Sep 2022 00:00:00 GMT 0.5 seconds average response time in logistics. Not a benchmark — a business requirement. https://thisisfatih.com/posts/european-patent-ep-4-323-939 https://thisisfatih.com/posts/european-patent-ep-4-323-939 Mon, 18 Jul 2022 00:00:00 GMT What the patent covers, why we filed, and what the process actually looks like from inside a startup. https://thisisfatih.com/posts/selling-ai-to-enterprise https://thisisfatih.com/posts/selling-ai-to-enterprise Mon, 09 May 2022 00:00:00 GMT Closed pilots with LVMH, Lego, Moncler, Puma, and Hugo Boss. What we got wrong before we got it right. https://thisisfatih.com/posts/building-ml-infrastructure-for-under-100k https://thisisfatih.com/posts/building-ml-infrastructure-for-under-100k Mon, 21 Mar 2022 00:00:00 GMT Enterprise computer vision at production scale does not require enterprise infrastructure budgets. What we built and what it cost. https://thisisfatih.com/posts/getting-to-99-point-6-percent-accuracy https://thisisfatih.com/posts/getting-to-99-point-6-percent-accuracy Mon, 10 Jan 2022 00:00:00 GMT In a warehouse, under fluorescent lights, with a conveyor belt at operational speed. What it took. https://thisisfatih.com/posts/classification-vs-detection-vs-segmentation https://thisisfatih.com/posts/classification-vs-detection-vs-segmentation Mon, 15 Nov 2021 00:00:00 GMT The wrong choice costs you six months. How to not make it. https://thisisfatih.com/posts/building-for-millions-of-requests https://thisisfatih.com/posts/building-for-millions-of-requests Mon, 20 Sep 2021 00:00:00 GMT When request volume stopped being a number we celebrated and started being a number we engineered for. https://thisisfatih.com/posts/computer-vision-in-logistics-specific-problems https://thisisfatih.com/posts/computer-vision-in-logistics-specific-problems Mon, 12 Jul 2021 00:00:00 GMT The papers show clean benchmark datasets. The logistics environment is not that. https://thisisfatih.com/posts/raising-the-seed-technical-founders-fundraising-notes https://thisisfatih.com/posts/raising-the-seed-technical-founders-fundraising-notes Mon, 24 May 2021 00:00:00 GMT What closing a seed round looks like from inside it, as a technical founder who had never raised before. https://thisisfatih.com/posts/why-we-chose-yolo-for-anti-counterfeiting https://thisisfatih.com/posts/why-we-chose-yolo-for-anti-counterfeiting Mon, 08 Mar 2021 00:00:00 GMT The architecture decision that comes up in every investor conversation, written down clearly once. https://thisisfatih.com/posts/founding-countercheck-first-90-days https://thisisfatih.com/posts/founding-countercheck-first-90-days Mon, 18 Jan 2021 00:00:00 GMT Three months in. What actually happened, before the story gets cleaned up in retrospect. https://thisisfatih.com/posts/tim-drapers-hero-mindset https://thisisfatih.com/posts/tim-drapers-hero-mindset Mon, 14 Dec 2020 00:00:00 GMT Two years watching the Hero Mindset framework applied by founders. What it means in practice. https://thisisfatih.com/posts/fifty-conversations-before-writing-a-line-of-code https://thisisfatih.com/posts/fifty-conversations-before-writing-a-line-of-code Mon, 05 Oct 2020 00:00:00 GMT How we validated Countercheck before building anything — and what 50 customer conversations actually produces. https://thisisfatih.com/posts/the-eir-model https://thisisfatih.com/posts/the-eir-model Mon, 10 Aug 2020 00:00:00 GMT Six months as an Entrepreneur in Residence. What the model actually is, from someone who's done it. https://thisisfatih.com/posts/gpt-3-and-the-moment-the-ceiling-got-removed https://thisisfatih.com/posts/gpt-3-and-the-moment-the-ceiling-got-removed Mon, 22 Jun 2020 00:00:00 GMT Early API access. Here's what it actually does. https://thisisfatih.com/posts/moving-draper-online-in-three-weeks https://thisisfatih.com/posts/moving-draper-online-in-three-weeks Mon, 06 Apr 2020 00:00:00 GMT COVID hit. Three weeks to move a residential global entrepreneurship program fully online without losing quality. https://thisisfatih.com/posts/anti-counterfeiting-as-a-computer-vision-problem https://thisisfatih.com/posts/anti-counterfeiting-as-a-computer-vision-problem Mon, 17 Feb 2020 00:00:00 GMT Mapping whether computer vision is the right fit for automating product authentication at scale in logistics and retail. https://thisisfatih.com/posts/the-decade-in-ai-2010-to-2019 https://thisisfatih.com/posts/the-decade-in-ai-2010-to-2019 Mon, 30 Dec 2019 00:00:00 GMT The hype arc of the 2010s, what actually changed, and what the predictions got badly wrong. https://thisisfatih.com/posts/what-zillion-pitches-taught-me-about-building-with-ai https://thisisfatih.com/posts/what-zillion-pitches-taught-me-about-building-with-ai Mon, 18 Nov 2019 00:00:00 GMT Two years, several hundred pitches analyzed. What we actually learned about building AI products. https://thisisfatih.com/posts/aaron-levie-on-building-box https://thisisfatih.com/posts/aaron-levie-on-building-box Mon, 07 Oct 2019 00:00:00 GMT Aaron Levie on why distribution is not a post-product problem — and how Box figured out enterprise go-to-market. https://thisisfatih.com/posts/why-most-ai-features-are-theater https://thisisfatih.com/posts/why-most-ai-features-are-theater Mon, 19 Aug 2019 00:00:00 GMT AI theater: when a product uses AI to perform intelligence rather than to produce it. https://thisisfatih.com/posts/tony-hsieh-on-culture https://thisisfatih.com/posts/tony-hsieh-on-culture Mon, 08 Jul 2019 00:00:00 GMT Culture is what you tolerate, not what you declare. Notes from Tony Hsieh's session with founders. https://thisisfatih.com/posts/what-phil-libin-talked-about https://thisisfatih.com/posts/what-phil-libin-talked-about Mon, 20 May 2019 00:00:00 GMT Phil Libin on asymmetric retention — why some products get more valuable the longer you use them. https://thisisfatih.com/posts/gpt-2-dropped-here-is-what-i-think-it-means https://thisisfatih.com/posts/gpt-2-dropped-here-is-what-i-think-it-means Wed, 03 Apr 2019 00:00:00 GMT OpenAI didn't release the full model. The stated reason was misuse. Here's what I think it actually means. https://thisisfatih.com/posts/speech-to-text-in-production https://thisisfatih.com/posts/speech-to-text-in-production Mon, 11 Feb 2019 00:00:00 GMT A year of running STT on founder pitch recordings. What the benchmark numbers don't tell you. https://thisisfatih.com/posts/nlp-in-2018-mostly-heuristics-with-a-transformer-on-top https://thisisfatih.com/posts/nlp-in-2018-mostly-heuristics-with-a-transformer-on-top Mon, 10 Dec 2018 00:00:00 GMT BERT dropped in October. What it changes for practitioners doing NLP in production. https://thisisfatih.com/posts/what-biz-stone-talked-about https://thisisfatih.com/posts/what-biz-stone-talked-about Mon, 15 Oct 2018 00:00:00 GMT Biz Stone on taste as a product skill — and why the discipline of removal is harder than addition. https://thisisfatih.com/posts/draper-university-inside-view https://thisisfatih.com/posts/draper-university-inside-view Mon, 03 Sep 2018 00:00:00 GMT Six weeks, 60 founders from 40 countries, and what actually changes people — from the program side. https://thisisfatih.com/posts/sentiment-analysis-in-production https://thisisfatih.com/posts/sentiment-analysis-in-production Mon, 09 Jul 2018 00:00:00 GMT The paper says 92% accuracy. The production system says something different. https://thisisfatih.com/posts/building-pitch-analysis-with-ai https://thisisfatih.com/posts/building-pitch-analysis-with-ai Mon, 14 May 2018 00:00:00 GMT Several hundred pitches analyzed. What NLP actually surfaces — and where the ceiling is. https://thisisfatih.com/posts/ibm-watson-the-gap-between-demo-and-integration https://thisisfatih.com/posts/ibm-watson-the-gap-between-demo-and-integration Mon, 05 Mar 2018 00:00:00 GMT Three months building on Watson for Zillion Pitches. The gap between the demo and the integration is real, and nobody talks about it. https://thisisfatih.com/posts/san-francisco-2018-what-the-city-taught-me https://thisisfatih.com/posts/san-francisco-2018-what-the-city-taught-me Mon, 22 Jan 2018 00:00:00 GMT The ambient conversations alone changed how I thought. You can't get that remotely. https://thisisfatih.com/posts/deep-learning-became-boring https://thisisfatih.com/posts/deep-learning-became-boring Mon, 27 Nov 2017 00:00:00 GMT When a technology becomes boring, it becomes useful. Notes on the maturation of the stack. https://thisisfatih.com/posts/why-im-going-to-san-francisco https://thisisfatih.com/posts/why-im-going-to-san-francisco Mon, 11 Sep 2017 00:00:00 GMT The pull of the AI wave. The decision to go be closer to where things are happening. https://thisisfatih.com/posts/one-year-as-first-time-cto https://thisisfatih.com/posts/one-year-as-first-time-cto Mon, 24 Jul 2017 00:00:00 GMT What I built, what I broke, and the one decision I'd reverse. https://thisisfatih.com/posts/expanding-to-the-us-with-too-much-confidence https://thisisfatih.com/posts/expanding-to-the-us-with-too-much-confidence Mon, 08 May 2017 00:00:00 GMT Turkey to Silicon Valley. What the ecosystem actually looks like when you're an outsider. https://thisisfatih.com/posts/the-gap-between-ai-powered-and-ai-that-works https://thisisfatih.com/posts/the-gap-between-ai-powered-and-ai-that-works Mon, 06 Mar 2017 00:00:00 GMT 2017. Every startup deck says AI. Almost none of them mean it. https://thisisfatih.com/posts/azure-cognitive-services-2017-honest-review https://thisisfatih.com/posts/azure-cognitive-services-2017-honest-review Mon, 16 Jan 2017 00:00:00 GMT We used them in production. Here's what the docs don't prepare you for. https://thisisfatih.com/posts/shipping-to-app-store-and-google-play-simultaneously https://thisisfatih.com/posts/shipping-to-app-store-and-google-play-simultaneously Mon, 21 Nov 2016 00:00:00 GMT Not a tutorial. A survival guide for doing both platforms at once with no dedicated mobile team. https://thisisfatih.com/posts/cto-at-mid-twenties https://thisisfatih.com/posts/cto-at-mid-twenties Mon, 05 Sep 2016 00:00:00 GMT The gap between writing code and leading technical direction is wider than it looks. https://thisisfatih.com/posts/microsoft-bizspark-what-120k-in-azure-credits-gets-you https://thisisfatih.com/posts/microsoft-bizspark-what-120k-in-azure-credits-gets-you Mon, 11 Jul 2016 00:00:00 GMT Free cloud infrastructure is a gift. It's also a trap if you're not careful. https://thisisfatih.com/posts/winning-y2fi-and-what-nobody-tells-you https://thisisfatih.com/posts/winning-y2fi-and-what-nobody-tells-you Sun, 22 May 2016 00:00:00 GMT We won and immediately started making expensive decisions. https://thisisfatih.com/posts/building-a-social-app-solo-xamarin https://thisisfatih.com/posts/building-a-social-app-solo-xamarin Mon, 14 Mar 2016 00:00:00 GMT One engineer. Two platforms. One backend. The decisions that kept me sane. https://thisisfatih.com/posts/tensorflow-0x-first-impressions https://thisisfatih.com/posts/tensorflow-0x-first-impressions Tue, 19 Jan 2016 00:00:00 GMT Google open-sourced it two months ago. I've been poking at it since. Here's where I am. https://thisisfatih.com/posts/the-year-i-stopped-being-afraid-of-machine-learning https://thisisfatih.com/posts/the-year-i-stopped-being-afraid-of-machine-learning Tue, 03 Nov 2015 00:00:00 GMT A note from someone who was convinced ML was for PhDs. It isn't. https://thisisfatih.com/posts/why-i-rewrote-my-msc-thesis-three-times https://thisisfatih.com/posts/why-i-rewrote-my-msc-thesis-three-times Mon, 14 Sep 2015 00:00:00 GMT Automated testing for mobile apps that generate sensory data. The problem kept reshaping itself. https://thisisfatih.com/posts/understanding-backpropagation-without-the-math-degree https://thisisfatih.com/posts/understanding-backpropagation-without-the-math-degree Wed, 08 Jul 2015 00:00:00 GMT I spent three weeks on this. Here's the version I wish someone had written for me. https://thisisfatih.com/posts/my-first-scikit-learn-classifier https://thisisfatih.com/posts/my-first-scikit-learn-classifier Tue, 12 May 2015 00:00:00 GMT Built a text classifier for my MSc thesis. Got 94% accuracy. Shipped garbage.