[{"content":" Agoston Fung Budapest, Hungary\n+36 20 338 7096\nagoston.fung@gmail.com\nlinkedin.com/in/agostonfung Engineer with 20 years of experience building and scaling technology solutions across corporate and startup environments. Proven track record in leading teams, delivering complex projects, and driving both technical and organisational growth. Experienced in architecture, cloud, agile delivery, and aligning technology with business goals. Equally comfortable as a hands-on technologist, strategic leader, or startup advisor. Experience AI/ML Engineer \u0026mdash; Ancestralize 2026 \u0026ndash; present Leading GenAI integration and agentic workflow development for medical researchers curing chronic diseases using a Bayesian network. Coaching teams to adopt agentic engineering practices.\nCTO, Co-founder \u0026mdash; RxClarity 2025 Bootstrapped a health-tech startup that provides enriched biomedical data for pharmaceutical companies using GenAI, RAG and agents.\nTechnical Lead \u0026mdash; EY 2024 \u0026ndash; 2025 Leading a team of senior engineers working on a master data management system integrated with core EY databases and a high performance tax reporting project for US enterprises.\nTechnical Lead \u0026mdash; bp 2023 \u0026ndash; 2024 Innovating bp's safety portfolio with a document store for relevant worldwide incident learning material enriched with metadata.\nEngineering Manager \u0026mdash; Tesco 2020 \u0026ndash; 2023 Leading teams to rewrite core legacy systems, prototype new solutions and innovate on Tesco's commercial income finance processes.\nEducation MSc Psychology (Distinction) \u0026mdash; Kingston University, UK 2013 Pursued my interest in the topic while working part-time.\nDissertation: Personality and coping with stress.\nMSc Computer Science \u0026mdash; BME, HU 2003 \u0026ndash; 2009 Interest in web development. Started working full time after my second year.\nDissertation: Prototype of a collaboration system for government use.\nProgramming C#JavaKotlinPythonTypeScript Passion AI/MLArchitecturePsych. safetyAgentic Eng.System Design Software Micro ServicesCost EffectiveComplianceIntegrationScalabilitySecurityDevopsCloudAPIsIaC Platforms WebMobileDesktop Hobbies Father of two sonsComputer GamesPhotographySwimmingSkiingI also like coffee ","permalink":"/cv/","summary":"\u003cdiv class=\"cv\"\u003e\n\u003cdiv class=\"cv-header\"\u003e\n  \u003ch1\u003eAgoston Fung\u003c/h1\u003e\n  \u003ca class=\"cv-download-icon\" href=\"/files/agoston-fung-cv.pdf\" download aria-label=\"Download CV (PDF)\" title=\"Download CV (PDF)\"\u003e\n    \u003csvg xmlns=\"http://www.w3.org/2000/svg\" viewBox=\"0 0 24 24\" fill=\"none\" stroke=\"currentColor\" stroke-width=\"2\" stroke-linecap=\"round\" stroke-linejoin=\"round\"\u003e\u003cpath d=\"M21 15v4a2 2 0 0 1-2 2H5a2 2 0 0 1-2-2v-4\"\u003e\u003c/path\u003e\u003cpolyline points=\"7 10 12 15 17 10\"\u003e\u003c/polyline\u003e\u003cline x1=\"12\" y1=\"15\" x2=\"12\" y2=\"3\"\u003e\u003c/line\u003e\u003c/svg\u003e\n  \u003c/a\u003e\n\u003c/div\u003e\n\u003cdiv class=\"cv-contact\"\u003e\n  Budapest, Hungary\u003cbr\u003e\n  +36 20 338 7096\u003cbr\u003e\n  \u003ca href=\"mailto:agoston.fung@gmail.com\"\u003eagoston.fung@gmail.com\u003c/a\u003e\u003cbr\u003e\n  \u003ca href=\"https://www.linkedin.com/in/agostonfung\"\u003elinkedin.com/in/agostonfung\u003c/a\u003e\n\u003c/div\u003e\n\u003cp class=\"cv-summary\"\u003e\nEngineer with 20 years of experience building and scaling technology\nsolutions across corporate and startup environments. Proven track record in\nleading teams, delivering complex projects, and driving both technical and\norganisational growth. Experienced in architecture, cloud, agile delivery,\nand aligning technology with business goals. Equally comfortable as a\nhands-on technologist, strategic leader, or startup advisor.\n\u003c/p\u003e","title":"CV"},{"content":"A few photos from travels over the years, alongside the engineering. All shot on trips around Europe and Asia; a wider set is up on Pexels. Click a photo to view it full screen.\nSeven Sisters, Seaford, England Winter lake, Hungary Krk, Croatia Lake Balaton, Hungary Terraced mountains, China Koh Samui, Thailand Angkor Wat, Cambodia Barcelona, Spain ","permalink":"/photography/","summary":"\u003cp\u003eA few photos from travels over the years, alongside the engineering.\nAll shot on trips around Europe and Asia; a wider set is up on\n\u003ca href=\"https://www.pexels.com/hu-hu/@agoston-fung-1165130/\"\u003ePexels\u003c/a\u003e. Click a\nphoto to view it full screen.\u003c/p\u003e\n\u003cdiv class=\"photo-grid\"\u003e\n\u003cdiv class=\"photo-grid-item\" role=\"button\" tabindex=\"0\" data-lightbox-open=\"lb-seven-sisters\"\u003e\n  \u003cfigure\u003e\n    \u003cimg src=\"seven-sisters-england.jpg\" alt=\"Chalk cliffs of the Seven Sisters, Seaford, England\"\u003e\n    \u003cfigcaption\u003eSeven Sisters, Seaford, England\u003c/figcaption\u003e\n  \u003c/figure\u003e\n\u003c/div\u003e\n\u003cdiv class=\"photo-grid-item\" role=\"button\" tabindex=\"0\" data-lightbox-open=\"lb-winter-lake\"\u003e\n  \u003cfigure\u003e\n    \u003cimg src=\"winter-lake-hungary.jpg\" alt=\"Frozen lake in winter, Hungary\"\u003e\n    \u003cfigcaption\u003eWinter lake, Hungary\u003c/figcaption\u003e\n  \u003c/figure\u003e\n\u003c/div\u003e\n\u003cdiv class=\"photo-grid-item\" role=\"button\" tabindex=\"0\" data-lightbox-open=\"lb-krk-beach\"\u003e\n  \u003cfigure\u003e\n    \u003cimg src=\"krk-beach-croatia.jpg\" alt=\"Beach at Vrbnik, Krk, Croatia\"\u003e\n    \u003cfigcaption\u003eKrk, Croatia\u003c/figcaption\u003e\n  \u003c/figure\u003e\n\u003c/div\u003e\n\u003cdiv class=\"photo-grid-item\" role=\"button\" tabindex=\"0\" data-lightbox-open=\"lb-balaton\"\u003e\n  \u003cfigure\u003e\n    \u003cimg src=\"balaton-hungary.jpg\" alt=\"Lake Balaton, Hungary\"\u003e\n    \u003cfigcaption\u003eLake Balaton, Hungary\u003c/figcaption\u003e\n  \u003c/figure\u003e\n\u003c/div\u003e\n\u003cdiv class=\"photo-grid-item\" role=\"button\" tabindex=\"0\" data-lightbox-open=\"lb-mountains\"\u003e\n  \u003cfigure\u003e\n    \u003cimg src=\"mountains-china.jpg\" alt=\"Green terraced mountains, China\"\u003e\n    \u003cfigcaption\u003eTerraced mountains, China\u003c/figcaption\u003e\n  \u003c/figure\u003e\n\u003c/div\u003e\n\u003cdiv class=\"photo-grid-item\" role=\"button\" tabindex=\"0\" data-lightbox-open=\"lb-koh-samui\"\u003e\n  \u003cfigure\u003e\n    \u003cimg src=\"koh-samui-thailand.jpg\" alt=\"Beach at Koh Samui, Thailand\"\u003e\n    \u003cfigcaption\u003eKoh Samui, Thailand\u003c/figcaption\u003e\n  \u003c/figure\u003e\n\u003c/div\u003e\n\u003cdiv class=\"photo-grid-item\" role=\"button\" tabindex=\"0\" data-lightbox-open=\"lb-angkor-wat\"\u003e\n  \u003cfigure\u003e\n    \u003cimg src=\"angkor-wat-cambodia.jpg\" alt=\"Angkor Wat temple, Siem Reap, Cambodia\"\u003e\n    \u003cfigcaption\u003eAngkor Wat, Cambodia\u003c/figcaption\u003e\n  \u003c/figure\u003e\n\u003c/div\u003e\n\u003cdiv class=\"photo-grid-item\" role=\"button\" tabindex=\"0\" data-lightbox-open=\"lb-barcelona\"\u003e\n  \u003cfigure\u003e\n    \u003cimg src=\"barcelona-spain.jpg\" alt=\"Seagull over the beach in Barcelona, Spain\"\u003e\n    \u003cfigcaption\u003eBarcelona, Spain\u003c/figcaption\u003e\n  \u003c/figure\u003e\n\u003c/div\u003e\n\u003c/div\u003e\n\u003cdiv id=\"lb-seven-sisters\" class=\"lightbox\" data-lightbox\u003e\n  \u003cbutton type=\"button\" class=\"lightbox-close\" data-lightbox-close aria-label=\"Close\"\u003e\u003c/button\u003e\n  \u003cimg src=\"seven-sisters-england.jpg\" alt=\"Chalk cliffs of the Seven Sisters, Seaford, England\"\u003e\n\u003c/div\u003e\n\u003cdiv id=\"lb-winter-lake\" class=\"lightbox\" data-lightbox\u003e\n  \u003cbutton type=\"button\" class=\"lightbox-close\" data-lightbox-close aria-label=\"Close\"\u003e\u003c/button\u003e\n  \u003cimg src=\"winter-lake-hungary.jpg\" alt=\"Frozen lake in winter, Hungary\"\u003e\n\u003c/div\u003e\n\u003cdiv id=\"lb-krk-beach\" class=\"lightbox\" data-lightbox\u003e\n  \u003cbutton type=\"button\" class=\"lightbox-close\" data-lightbox-close aria-label=\"Close\"\u003e\u003c/button\u003e\n  \u003cimg src=\"krk-beach-croatia.jpg\" alt=\"Beach at Vrbnik, Krk, Croatia\"\u003e\n\u003c/div\u003e\n\u003cdiv id=\"lb-balaton\" class=\"lightbox\" data-lightbox\u003e\n  \u003cbutton type=\"button\" class=\"lightbox-close\" data-lightbox-close aria-label=\"Close\"\u003e\u003c/button\u003e\n  \u003cimg src=\"balaton-hungary.jpg\" alt=\"Lake Balaton, Hungary\"\u003e\n\u003c/div\u003e\n\u003cdiv id=\"lb-mountains\" class=\"lightbox\" data-lightbox\u003e\n  \u003cbutton type=\"button\" class=\"lightbox-close\" data-lightbox-close aria-label=\"Close\"\u003e\u003c/button\u003e\n  \u003cimg src=\"mountains-china.jpg\" alt=\"Green terraced mountains, China\"\u003e\n\u003c/div\u003e\n\u003cdiv id=\"lb-koh-samui\" class=\"lightbox\" data-lightbox\u003e\n  \u003cbutton type=\"button\" class=\"lightbox-close\" data-lightbox-close aria-label=\"Close\"\u003e\u003c/button\u003e\n  \u003cimg src=\"koh-samui-thailand.jpg\" alt=\"Beach at Koh Samui, Thailand\"\u003e\n\u003c/div\u003e\n\u003cdiv id=\"lb-angkor-wat\" class=\"lightbox\" data-lightbox\u003e\n  \u003cbutton type=\"button\" class=\"lightbox-close\" data-lightbox-close aria-label=\"Close\"\u003e\u003c/button\u003e\n  \u003cimg src=\"angkor-wat-cambodia.jpg\" alt=\"Angkor Wat temple, Siem Reap, Cambodia\"\u003e\n\u003c/div\u003e\n\u003cdiv id=\"lb-barcelona\" class=\"lightbox\" data-lightbox\u003e\n  \u003cbutton type=\"button\" class=\"lightbox-close\" data-lightbox-close aria-label=\"Close\"\u003e\u003c/button\u003e\n  \u003cimg src=\"barcelona-spain.jpg\" alt=\"Seagull over the beach in Barcelona, Spain\"\u003e\n\u003c/div\u003e\n\u003cscript\u003e\n(function () {\n  function close(box) {\n    box.classList.remove('is-open');\n  }\n  document.querySelectorAll('[data-lightbox-open]').forEach(function (item) {\n    function open() {\n      var box = document.getElementById(item.getAttribute('data-lightbox-open'));\n      if (box) box.classList.add('is-open');\n    }\n    item.addEventListener('click', open);\n    item.addEventListener('keydown', function (e) {\n      if (e.key === 'Enter' || e.key === ' ') {\n        e.preventDefault();\n        open();\n      }\n    });\n  });\n  document.querySelectorAll('[data-lightbox]').forEach(function (box) {\n    box.addEventListener('click', function (e) {\n      if (e.target === box || e.target.hasAttribute('data-lightbox-close')) {\n        close(box);\n      }\n    });\n  });\n  document.addEventListener('keydown', function (e) {\n    if (e.key === 'Escape') {\n      document.querySelectorAll('.lightbox.is-open').forEach(close);\n    }\n  });\n})();\n\u003c/script\u003e","title":"Photography"},{"content":"In my previous startup, I jumped on the AI coding train and fully immersed myself in running agents instead of writing code. 🤖 At first it felt incredible. I was 10x more productive, doing in a fraction of the time all the things I\u0026rsquo;d wanted to do. Hard changes became easy, and I was producing more code in days than I used to produce in months. 🚀\nBut at some point, I broke. I\u0026rsquo;d sometimes spend one or two days just lying in bed, mentally exhausted, needing to do anything except work. I\u0026rsquo;d take days off mid-week because I simply couldn\u0026rsquo;t keep going. This went on for weeks, until I needed an entire week to recover. Eventually I couldn\u0026rsquo;t even look at the code anymore - I\u0026rsquo;d start my computer, open my IDE, look at the tasks in front of me, and still couldn\u0026rsquo;t get my brain to focus. I needed a longer break, 4 to 6 weeks, before I was ready to work again.\nAI fatigue is real. More people are starting to talk about it. 🧠\nWhat I took away from it:\nI can spend about 3 to 4 hours a day prompting agents. The rest of the day is needed for recovery and for my brain to catch up. I have to force myself not to work in between. My brain keeps processing everything anyway, so that time isn\u0026rsquo;t wasted. Taking care of myself matters more now than ever: regular exercise, walks in the park, healthy food, limiting caffeine ☕, finding ways to relax, and spending time with people - real conversations, with people who are physically there. I can push past these limits for a while, but there\u0026rsquo;s no free lunch. Doing more today eventually means doing less tomorrow, or worse. Have you experienced AI fatigue yet? 👇\nOriginally posted on LinkedIn.\n","permalink":"/posts/ai-fatigue-is-real/","summary":"Running agents instead of writing code felt like 10x productivity. Then I broke, and needed 4-6 weeks to recover.","title":"AI fatigue is real"},{"content":"GUI is dead. Long live agents. 🤖\nComputer interfaces keep evolving: command line → thick clients → the web → and now, agents. Most people move through the same stages with agentic engineering:\nLevel 1: \u0026ldquo;I\u0026rsquo;m getting so much more done.\u0026rdquo; ⚡ AI speeds up how you write code.\nLevel 2: \u0026ldquo;I can build all these features myself, ship to every platform.\u0026rdquo; 🚀 AI removes the limits on what you can scope and build.\nLevel 3: \u0026ldquo;Why build a UI at all? I can expose the data via MCP and let users\u0026rsquo; own agents pull data, generate reports, take action directly.\u0026rdquo; 🔌 The interface itself becomes optional.\nThe front end is dissolving into the protocol layer. The product isn\u0026rsquo;t the screen anymore, it\u0026rsquo;s the API your users\u0026rsquo; agents can talk to. 👀\nOriginally posted on LinkedIn.\n","permalink":"/posts/gui-is-dead-long-live-agents/","summary":"Command line, thick clients, the web, and now agents. The interface itself is starting to look optional.","title":"GUI is dead, long live agents"},{"content":"The application needs to be rewritten. Better tech, better architecture, better code. Management says no. 🙅\nThe reason is always the same: big-bang migrations rarely work, and engineers spend weeks or months with nothing visible to show the business. So you don\u0026rsquo;t rewrite it, you improve it step by step. That\u0026rsquo;s been best practice for decades.\nUntil AI agents showed up. 🤖\nNow an agent can review tens of thousands of lines in seconds, absorb the docs for any technology, and rewrite an app in hours instead of months. Zero risk, 100% gain. Not everyone has realised this yet.\nYour project would benefit from an iOS app you\u0026rsquo;ve never had time to build? 💥 Done. Your backend is PHP 8 or Java 7, and a consultancy quoted $200k to move it to TypeScript? 💥 Done. Your app is still jQuery and static HTML? Pick your stack. 💥 Done. It\u0026rsquo;s time to modernise your tech stack and reap the benefits. 🚀\nOriginally posted on LinkedIn.\n","permalink":"/posts/ai-can-rewrite-your-legacy-app-in-hours/","summary":"Big-bang rewrites used to be a bad idea for good reasons. AI agents just changed the math.","title":"AI can rewrite your legacy app in hours"},{"content":"AI is supposed to make us more productive. But is it, really? 🤔\nOnce you know something, there\u0026rsquo;s no going back. The burden it puts on you has to be handled, and that\u0026rsquo;s energy on its own. Sometimes blissful ignorance helps you in ways you don\u0026rsquo;t realize.\nTwo examples:\n1️⃣ I asked my coding agent to review the security of my app. I just wanted to make sure there was no easy way to get hacked, maybe even automatically by a bot. Now all I can think about are the 10 security holes it found. Four are \u0026ldquo;low severity,\u0026rdquo; sure, but now I know. And if we get hacked, I can\u0026rsquo;t tell my business partner I had no idea. A harmless question created a dilemma I can\u0026rsquo;t escape.\n2️⃣ I asked ChatGPT a legal question out of pure curiosity. That night I couldn\u0026rsquo;t sleep, convinced I was going to get sued. I ended up paying a lawyer to review my case and tell me, out loud, that there was nothing to worry about. I normally love that it\u0026rsquo;s thorough and precise - this time it cost me money, time, and a lot of energy solving a problem I\u0026rsquo;d created for myself. 😅\nJust because AI can hand you all the knowledge in the world in seconds doesn\u0026rsquo;t mean you should burden yourself with unlimited information. ⚖️\nWhere do you draw the line? 👇\nOriginally posted on LinkedIn.\n","permalink":"/posts/the-burden-of-knowledge/","summary":"AI is supposed to make us more productive. Once you know something, though, there\u0026rsquo;s no going back - and that has a cost of its own.","title":"The unexpected side effect of LLMs: the burden of knowledge"},{"content":"AI writes code now, so technically almost anything can be built quickly and efficiently. Engineers will all be laid off soon 😄\nSo I tried vibe coding a little Sudoku app myself. How hard can it be? A 9x9 grid and 3 rules, sounds easy, right? And honestly, I was right about the vibe coding part - I shipped a fully functioning app without ever looking at the code. 👀\nWhat surprised me was all the tiny details and hard decisions that showed up along the way:\nScoring system: it has to feel fair, reward both speed and precision, scale well, and still stay motivating. 🎯 Visual helpers: finding the balance between helping the player and solving the puzzle for them. Too few and the app feels clunky, too many and you remove the joy of solving it. 🧩 Difficulty levels: there are 10+ Sudoku solving techniques. Difficulty isn\u0026rsquo;t just how many numbers are missing, it\u0026rsquo;s how the puzzle has to be solved. One step too hard and the puzzle feels impossible; too easy everywhere and it gets repetitive. 🧠 Winning conditions, error highlighting, hints, drafts, undo, save and resume, statistics, settings, animations, etc. Then I showed it to five people and immediately got ten pieces of feedback, many of them directly contradicting each other. 😂\nAt that point it became obvious: AI can write all the code, but it still can\u0026rsquo;t magically create a great product. Anyway, the app finally made it to the App Store, so I can cross this one off my bucket list. ✅\nOriginally posted on LinkedIn.\n","permalink":"/posts/i-vibe-coded-a-sudoku-app/","summary":"AI can write all the code. It still can\u0026rsquo;t magically create a great product - the taste has to come from somewhere else.","title":"I vibe-coded a Sudoku app and learned that product taste doesn't come free"},{"content":"Should your AI agent have direct access to the production database? Unrestricted access is dangerous. Locking it down so hard the agent can\u0026rsquo;t do anything useful defeats the purpose. The answer is architectural guardrails, not a binary yes or no.\nSeven strategies that work:\nDedicated database user or role with read-only permissions on selected tables only. Read-only transactions - native in Postgres, and a cheap safety net. Separate connection pools, so agent traffic can never starve or block application traffic. Query timeouts, so a bad query from an agent can\u0026rsquo;t take down performance for everyone else. Staging tables for writes - the agent writes there, a review step promotes changes to production. A dedicated read-only schema, scoped down to exactly what the agent needs to see. Read replica access instead of the primary, so exploration never touches the system of record. Most teams are treating AI agents like untrusted interns. The teams getting the best results are treating them like highly capable analysts, operating inside carefully designed guardrails.\nThe goal isn\u0026rsquo;t maximum restriction, it\u0026rsquo;s the right restriction. 🔒\nOriginally posted on LinkedIn.\n","permalink":"/posts/seven-strategies-for-safe-ai-agent-database-access/","summary":"Unrestricted database access for AI agents is dangerous. Excessive restrictions defeat the purpose. Here\u0026rsquo;s how to find the middle ground.","title":"Seven strategies for safe AI agent database access"},{"content":"Being a software engineer is weird when it comes to family. Many of us feel misunderstood, and honestly, it makes sense - what we do is rarely easy to grasp for a general audience.\nHave you had this at a family dinner? Everyone talks about their life, your turn comes, someone asks what you\u0026rsquo;ve been working on. You explain, short or long. The reaction is always the same: dead silence, then that subtle shift to change the topic. \u0026ldquo;Let\u0026rsquo;s open presents!\u0026rdquo; 🎁\nThen I ask my agent to \u0026ldquo;write me a Sudoku app,\u0026rdquo; show it around the table, and people lose their minds. 🤯 \u0026ldquo;Oh wow, you can do that?\u0026rdquo; I have never received this much attention, compliments, and validation for my skills as I did after three days of vibe coding a Sudoku app.\nFor comparison:\nMy team and I spent years automating the invoicing of £50 million in promotional income from 20 million sales events a week. → silence My team and I built a system handling tax reporting for 60 million people. → silence Here\u0026rsquo;s a Sudoku app I asked my agent to write. → admiration, excitement, basically a star on the Hollywood Walk of Fame ⭐ If you want to take advantage of this moment too, the window might not stay open long. Agents are getting more powerful, and more expensive, fast. ⏳\nOriginally posted on LinkedIn.\n","permalink":"/posts/being-a-software-engineer-is-weird-at-family-dinners/","summary":"My team spent years automating £50 million in promotional income. A three-day Sudoku app got me a standing ovation at dinner.","title":"Being a software engineer is weird when it comes to family"},{"content":"This is a lesson I recently learned in business: pressure doesn\u0026rsquo;t create character, it reveals it.\nUnder stress, people tend to fall into one of two patterns:\nThey communicate openly, reach out to others, and invite collaboration in solving the problem. They get defensive, deflect blame, or quietly shift responsibility elsewhere. The same split shows up in how people handle failure. Some take ownership and focus on what to improve next. Others externalise it, and the story becomes about anything except their part in it.\nThese aren\u0026rsquo;t fixed personality types, more like defaults people fall back on when things get hard. But the patterns are consistent enough that pressure becomes a genuinely fast way to learn who you\u0026rsquo;re working with.\nWorth being intentional about it: work with people who communicate, take ownership, and treat others with respect, especially when things aren\u0026rsquo;t going well. That last part matters most exactly when it\u0026rsquo;s hardest to do.\nOriginally posted on LinkedIn.\n","permalink":"/posts/pressure-reveals-patterns/","summary":"Under stress, people fall into a handful of default patterns - and pressure is the fastest way to find out who you\u0026rsquo;re actually working with.","title":"Pressure reveals patterns"},{"content":"I\u0026rsquo;m Agoston Fung, a software engineer based in Budapest, Hungary, with 20+ years in the industry. My career has spanned startups and large enterprises — including EY, bp, and Tesco — and I currently consult through Equal Experts.\nI like working at the intersection of clean code, pragmatic systems design, and teams that ship. This site is where I think out loud about that work.\nBackground Currently: Consultant, Equal Experts Previously: EY, bp, Tesco, and a handful of startups Education: Kingston University, London (Distinction) Languages: Hungarian (native), English (professional) Selected publications Different coping strategies amongst individuals with grandiose and vulnerable narcissistic traits — Journal of Affective Disorders, 2016 Introduction to the Symfony framework (Hungarian) — 2007 Elsewhere LinkedIn GitHub Email The bio above is intentionally short — feel free to expand it with specific roles, dates, and projects you\u0026rsquo;d like to highlight.\n","permalink":"/about/","summary":"\u003cp\u003eI\u0026rsquo;m \u003cstrong\u003eAgoston Fung\u003c/strong\u003e, a software engineer based in Budapest, Hungary, with\n20+ years in the industry. My career has spanned startups and large\nenterprises — including EY, bp, and Tesco — and I currently consult through\n\u003ca href=\"https://www.equalexperts.com/\"\u003eEqual Experts\u003c/a\u003e.\u003c/p\u003e\n\u003cp\u003eI like working at the intersection of clean code, pragmatic systems design,\nand teams that ship. This site is where I think out loud about that work.\u003c/p\u003e\n\u003ch2 id=\"background\"\u003eBackground\u003c/h2\u003e\n\u003cul\u003e\n\u003cli\u003e\u003cstrong\u003eCurrently:\u003c/strong\u003e Consultant, \u003ca href=\"https://www.equalexperts.com/\"\u003eEqual Experts\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003e\u003cstrong\u003ePreviously:\u003c/strong\u003e EY, bp, Tesco, and a handful of startups\u003c/li\u003e\n\u003cli\u003e\u003cstrong\u003eEducation:\u003c/strong\u003e Kingston University, London (Distinction)\u003c/li\u003e\n\u003cli\u003e\u003cstrong\u003eLanguages:\u003c/strong\u003e Hungarian (native), English (professional)\u003c/li\u003e\n\u003c/ul\u003e\n\u003ch2 id=\"selected-publications\"\u003eSelected publications\u003c/h2\u003e\n\u003cul\u003e\n\u003cli\u003e\u003cem\u003eDifferent coping strategies amongst individuals with grandiose and\nvulnerable narcissistic traits\u003c/em\u003e — Journal of Affective Disorders, 2016\u003c/li\u003e\n\u003cli\u003e\u003cem\u003eIntroduction to the Symfony framework\u003c/em\u003e (Hungarian) — 2007\u003c/li\u003e\n\u003c/ul\u003e\n\u003ch2 id=\"elsewhere\"\u003eElsewhere\u003c/h2\u003e\n\u003cul\u003e\n\u003cli\u003e\u003ca href=\"https://www.linkedin.com/in/agostonfung/\"\u003eLinkedIn\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003e\u003ca href=\"https://github.com/eshton\"\u003eGitHub\u003c/a\u003e\u003c/li\u003e\n\u003cli\u003e\u003ca href=\"mailto:agoston.fung@gmail.com\"\u003eEmail\u003c/a\u003e\u003c/li\u003e\n\u003c/ul\u003e\n\u003cblockquote\u003e\n\u003cp\u003eThe bio above is intentionally short — feel free to expand it with\nspecific roles, dates, and projects you\u0026rsquo;d like to highlight.\u003c/p\u003e","title":"About"},{"content":"The real cost of AI writing code is starting to show 👀\nFor the past 6 months, we\u0026rsquo;ve been coding like there\u0026rsquo;s no tomorrow: endless tokens, multiple agents running in parallel, shipping features in stacks and technologies we barely touched before 🚀\nBut the golden phase is fading. The signs are already here:\nAnthropic dropped their $20 subscription Token usage is getting tighter Companies are reporting hundreds of thousands to millions in monthly AI spend 💸 The pendulum might be swinging back. Smaller companies are starting to hire juniors again to write code. Larger companies are seeing senior engineers spend most of their time reviewing AI-generated output instead of building.\nWe\u0026rsquo;re not at equilibrium yet. That balance between humans and AI is still ahead of us ⚖️\nOriginally posted on LinkedIn.\n","permalink":"/posts/the-ai-honeymoon-is-over/","summary":"The real cost of AI writing code is starting to show - what the data is telling us about the human/AI balance in software work.","title":"The AI honeymoon is over"},{"content":"How can an LLM write code? 🤔 How can it come up with ideas? 💡 How can it summarize long documents? 📄\nIt can feel like AI is \u0026ldquo;thinking\u0026rdquo;. But what is actually happening is closer to pattern recognition.\nAn LLM is like someone who has read millions of crime stories. When you start telling a new one, they can quickly guess what usually happens next. Not because they know the truth, but because they have seen similar patterns many times. 🕵️‍♂️\nThe same idea applies to everything else:\n💻 Writing code - It has seen millions of code examples and learns common patterns\n💡 Coming up with ideas - It has seen brainstorming sessions, strategy docs, and idea lists\n📄 Summarizing text - It has seen many examples of long texts paired with shorter summaries\nSo instead of thinking from scratch, an LLM predicts what a good next response usually looks like, based on patterns learned from a huge amount of text.\nIn a way, LLMs are extremely powerful pattern recognition engines that learned from human knowledge at scale. 🌍\nWhat analogy do you use to explain AI to non technical people?\nOriginally posted on LinkedIn.\n","permalink":"/posts/how-can-an-llm-write-code/","summary":"It can feel like AI is thinking. What\u0026rsquo;s actually happening is closer to pattern recognition - and that frame makes LLMs much easier to reason about.","title":"How can an LLM write code?"},{"content":"Have you heard of the latest mental health phenomenon? Weirdly, it only affects programmers. Hence the name: Claude separation anxiety. 😰💻\nDefinition: when you have not seen actual code for such a long time that you forgot what it even looks like. All code now materializes through Claude. Your job is mostly nodding, copying, and saying \u0026ldquo;looks good\u0026rdquo;. 🤖✨\nSymptoms:\nMild panic when Claude says \u0026ldquo;I may be mistaken\u0026rdquo; 😬 Inability to write a for loop without emotional support 🫠 Reviewing code like a museum visitor observing ancient artifacts 🏛️👨‍💻 Side effects: You wake up at 3:17 AM in a cold sweat after a nightmare that the $90 subscription is now $9000. 😱\nYou check your email. Still $90. 🙏\nYou whisper thank you and go back to sleep.\nOriginally posted on LinkedIn.\n","permalink":"/posts/claude-separation-anxiety/","summary":"A small clinical case study of a new programming-related condition. Symptoms may include reviewing code like a museum visitor observing ancient artifacts.","title":"Claude separation anxiety"},{"content":"On a busy week, I make around 10,000 code line changes a day on average. 🤯\nAm I expected to review all that code? Pf. That is so last year thinking. Or maybe better said, 6 months ago thinking.\nNo. What I do instead is this:\nI make sure the architecture is clean, efficient, and designed for the problem I am solving. How? I ask the Agent to discuss options, pros and cons, and then I review it carefully to choose the architecture I want 🧠 Then I ask the Agent to code while I watch an episode of Friends ☕ Then I ask it to review the code and check for mistakes. It always finds 2–3 issues 🔍 Then I test the application manually, and report issues back to the Agent to fix Then I ask the Agent to write tests covering the main use cases, happy path, and the error scenarios I discovered 🧪 Then I tell the Agent: make the code production ready. Simple as that 🚀 Optionally, I ask it to focus separately on error handling, security, performance, and observability Finally, I ask it to run build, linting, and assess the code coverage I don\u0026rsquo;t need to review the code anymore. I review the thinking instead. 💡\nOriginally posted on LinkedIn.\n","permalink":"/posts/i-review-the-thinking-not-the-code/","summary":"On a busy week, I ship around 10,000 lines of code changes a day. There is no way to review all that line by line — and that\u0026rsquo;s not what review should be anymore.","title":"I don't review the code anymore. I review the thinking."},{"content":"Sam Altman once said millions of dollars are wasted on people thanking LLMs.\nSo people conclude you should not do it. It is pointless. Maybe even wrong.\nI don\u0026rsquo;t think so.\nWhen you talk to an LLM, psychologically you are not talking to a machine. You are talking to a mirror. Your brain still runs the \u0026ldquo;conversation with another human\u0026rdquo; mode. 🧠\nAnd that matters.\nTreat the LLM like shit and technically nothing happens. But something does change.\nYou.\nThe way you talk to an LLM is basically practice for how you talk in general. To anyone. Do it 100 times a day and your brain optimizes it. It gets cached. Autocomplete for your personality. ⚡\nAt some point you will slip and talk to a person the same way.\nKeep being respectful and the worst thing that happens is you become a slightly better human. Horrible outcome. 😄\nSo yes, thanking the LLM might waste a few GPU cycles 🤖\nBut being thoughtful in how you write? That investment compounds.\nOriginally posted on LinkedIn.\n","permalink":"/posts/autocomplete-for-your-personality/","summary":"Sam Altman said millions are wasted on people thanking LLMs. I keep doing it anyway. Here\u0026rsquo;s why - and what your brain does when you don\u0026rsquo;t.","title":"Autocomplete for your personality"},{"content":"Less than a year ago, a dear friend of mine (Zsolt Pocsaji) told me that very soon, software development wouldn\u0026rsquo;t need people to write code. Humans would only write code for true innovation. Everything else? Supervised AI agents doing the heavy lifting.\nAt first, I laughed. Internally… but still.\nThen I looked at his completely serious face and realized this wasn\u0026rsquo;t another Elon-level \u0026ldquo;Cybertruck is indestructible\u0026rdquo; moment. 😄\nThe timeline since that conversation:\n3 months: My agency rebrands as an AI-first consultancy 🤖\n5 months: I\u0026rsquo;m leading AI transformation across multiple teams… with heavy resistance 🧱\n8 months: I personally review \u0026lt;20% of the code my agent writes, while the internet is still debating \u0026ldquo;vibe coding\u0026rdquo; like it\u0026rsquo;s a philosophical movement 🧘‍♂️\n12 months: I\u0026rsquo;m building software almost entirely through prompts in Claude. The industry quietly accepts the new reality. Job titles evolve from \u0026ldquo;Vibe coding cleanup specialist\u0026rdquo; to simply… \u0026ldquo;AI engineer.\u0026rdquo; 🚀\nThe speed of change is enormous.\nFor comparison: when Ryan Dahl introduced Node.js in 2009, I told colleagues that one day everything would be written in JavaScript. They laughed. Loudly. Publicly. Probably still laughing somewhere. 😅\nAnd yet, JavaScript still needed 8 years to become the #1 programming language in the world.\nSo either: my friend is a genius (he is, obviously), I\u0026rsquo;m terrible at predicting the future, or… we\u0026rsquo;re living through truly unprecedented times.\nSlightly exciting. Slightly scary. Mostly both.\nOriginally posted on LinkedIn.\n","permalink":"/posts/less-than-a-year-ago/","summary":"A friend told me software wouldn\u0026rsquo;t need humans to write code. I laughed. Twelve months later, here\u0026rsquo;s the receipt.","title":"Less than a year ago"},{"content":"Do you want to be a better vibe-coder? 😎\nI have news: you need to learn to be a really bad boss. 😈\nIn the age of AI and Copilot, if you want to fully exploit all the new technologies, you also need to master some fundamental toxic management skills. 🤝\nSee, horrible bosses are great at:\nLevel 1:\n🔍 Micromanaging your every action 🙅 Thinking you cannot do anything good enough without them 🔁 Double, triple, quadruple checking your work Level 2:\n❌ Criticise everything you do 🚫 Do not allow you to make any decisions 📈 Demand results, question the impossible 🎭 Manipulate you into doing something you don\u0026rsquo;t want Ultimate level:\n🔥 Fire you as soon as a cheaper or better model is available 🧠 Reset your memory from time to time (er…) Well, it turns out LLM coding tools work best if you practice the virtues above. 🤖✨\nAnd most importantly, do not forget to take full credit for all the hard work your LLM has done for you. 😌\nNo one needs to know your secret recipe. 🤫\nThank me later. You are welcome! 😄\nOriginally posted on LinkedIn.\n","permalink":"/posts/to-be-a-better-vibe-coder-become-a-horrible-boss/","summary":"Want to fully exploit AI coding tools? Time to dust off your worst-boss instincts. A short field guide.","title":"To be a better vibe-coder, become a horrible boss"},{"content":"I couldn\u0026rsquo;t believe it either 😅\nWhen I first heard on Gergely Orosz\u0026rsquo;s podcast that some days Steve Yegge would change 50,000 lines of code with full vibe coding, I laughed. Impossible, right?\nSo I checked… and I was amazed.\nNearly 300,000 lines of code changed (added or removed) in my GitHub over the last 10 weeks. And I wasn\u0026rsquo;t even trying 😄\nA few caveats:\nI was prototyping 5 different services One project I deleted and rewrote from scratch with a better prompt 3 times 🤷‍♂️ But still, it\u0026rsquo;s obviously not 300,000 lines of production quality code. Around 50-70% is throwaway code… and honestly, that\u0026rsquo;s exactly what I needed.\nWhen I don\u0026rsquo;t know what I want: I just test it out! When I don\u0026rsquo;t know if the solution will work at all: I test it out! When I don\u0026rsquo;t know if the architecture makes sense: I test it out! No more analysis paralysis 🚀\nBuild, rebuild, and move in small steps toward real value.\nOriginally posted on LinkedIn.\n","permalink":"/posts/my-ai-productivity-4200-lines-per-day/","summary":"Nearly 300,000 lines of code changed in my GitHub over 10 weeks. Most of it is throwaway. That\u0026rsquo;s exactly the point.","title":"My AI productivity: 4,200 lines of code per day"},{"content":"Three categories I keep reaching for lately 🚀\nAzure / AWS ☁️ Connect straight to S3 or Azure Storage and query files directly Use secure, modern auth like RBAC, tokens and CLI creds JSON / CSV 📄 Load and explore semi-structured or flat files from anywhere Automatic schema inference that feels like magic Use asterisk wildcard to easily load multiple files Postgres / MySQL 🗄️ Attach external databases as if they were part of DuckDB Join and query across multiple systems without friction Stream data both ways for fast ETL and rapid prototyping Still not on the \u0026ldquo;ducktrain\u0026rdquo;? Have a look at duckdb.org.\nOriginally posted on LinkedIn.\n","permalink":"/posts/my-favourite-duckdb-extensions/","summary":"Three categories of DuckDB extensions I keep reaching for: cloud storage, semi-structured files, and external SQL databases.","title":"My favourite DuckDB extensions"},{"content":"Ever get stuck on a problem and think the solution is to grind harder? Lock yourself in a room, stare at the screen, and power through?\nYeah… no. Just imagining that makes my eyes water 😢.\nIt\u0026rsquo;s not only miserable - it\u0026rsquo;s actively worse for solving the problem.\nHere\u0026rsquo;s the counter-intuitive truth: walk away. 🚶‍♂️💨\nLiterally leave the task and come back later. And the tougher the problem, the more it benefits from that break.\nBut Agoston, this sounds like something you made up.\nNope. This is one of the most robust findings in creativity and problem-solving research: the incubation effect - stepping away improves solutions 🧠✨.\nBut Agoston, you haven\u0026rsquo;t been in uni for a decade.\nFair. So here are two solid studies from the last 10 years:\nHenok et al., 2018 - Incubation and interactivity in insight problem solving Gilhooly et al., 2016 - Incubation and Intuition in Creative Problem Solving Don\u0026rsquo;t be scared by the word incubation - it\u0026rsquo;s science-speak for \u0026ldquo;leave the problem alone for a bit\u0026rdquo; 🌱.\nAnd if you want more studies (there are plenty), just ask ChatGPT 🤖.\nOriginally posted on LinkedIn.\n","permalink":"/posts/stuck-on-a-problem-walk-away/","summary":"The counter-intuitive way to solve hard problems isn\u0026rsquo;t to grind harder. It\u0026rsquo;s to leave them alone. Here\u0026rsquo;s the research backing it - the incubation effect.","title":"Stuck on a problem? Walk away."},{"content":"To be an engineer, you need at least a BSc and an MSc in Computer Science. ❌ Wrong.\nIf you have imposter syndrome, you\u0026rsquo;re not cut out for engineering. ❌ Wrong.\nIf you burn out and take a career break, it means you\u0026rsquo;re not meant for this field. ❌ Wrong.\nIf you\u0026rsquo;re 30 and still figuring out whether coding is for you, it\u0026rsquo;s probably not. ❌ Wrong.\nIf you sometimes hate coding, you should quit. ❌ Wrong.\nThese are myths I\u0026rsquo;ve come across - and wrestled with - over 20 years in software development. 🧑‍💻\nHere\u0026rsquo;s what\u0026rsquo;s actually true:\n❤️ Passion and frustration come as a pair. You can love your craft and still have days you want to throw your laptop out the window.\n🎢 Ambition comes with doubt. Feeling like an imposter doesn\u0026rsquo;t mean you\u0026rsquo;re unfit - it often means you\u0026rsquo;re growing.\n🌱 Breaks are healthy. Trying something new, resting, exploring another path - none of these disqualify you from being an engineer.\nIf you\u0026rsquo;re struggling with any of these feelings, you\u0026rsquo;re not alone. 💬 Send me a message - happy to chat and share perspective.\nOriginally posted on LinkedIn.\n","permalink":"/posts/to-be-an-engineer-you-need-at-least-a-bsc/","summary":"Five myths I\u0026rsquo;ve wrestled with over 20 years in software - degrees, imposter syndrome, burnout, career breaks, and the days you hate coding. None of them are true.","title":"To be an engineer, you need at least a BSc"},{"content":"My memory is basically that of a goldfish. I only remember what truly matters - and SQL syntax does not make the cut.\nThe other day I needed to change a column type in my database, so I wrote something like:\nALTER TABLE rxc.trial change column drugs to text instead of varchar; Yes, I know. There are mistakes. Parts of it are basically fan-fiction.\nAnd at this point, some of you are already judging me. You probably call yourself a \u0026ldquo;data engineer\u0026rdquo; or an \u0026ldquo;SQL magician.\u0026rdquo; You wake up at 3 AM reciting the differences between Postgres 10 and 11. You speak SQL better than your native language.\nI am not you.\nI wrote that line purely so I could add this right under it:\n//TODO: fix the syntax of this statement Then I asked Copilot to \u0026ldquo;fix the TODO in this file.\u0026rdquo; Because I\u0026rsquo;m not just lazy - I\u0026rsquo;m committed to not learning. Goldfish, remember?\nBut here\u0026rsquo;s the thing: Copilot didn\u0026rsquo;t warn me that this change could cause downtime or even cost the company $50k. (It didn\u0026rsquo;t happen - but it could have.) And that\u0026rsquo;s exactly why I still need the SQL magician. They would have told me. Right after making fun of me for the syntax.\nSo no, AI didn\u0026rsquo;t replace you. It liberated you - from people like me - so you can focus on real problems instead of listening to my existential crisis about SQL.\nOriginally posted on LinkedIn.\n","permalink":"/posts/my-memory-is-basically-that-of-a-goldfish/","summary":"On SQL syntax, fan-fiction ALTER TABLE statements, why I still need the SQL magician, and what AI actually liberated us from.","title":"My memory is basically that of a goldfish"},{"content":"AI has already replaced junior and mid-level developers. 🚫\nWrong.\nThere has never been a better time to start a career in software. Here\u0026rsquo;s why 👇\n📚 Resources have never been better - courses, tutorials, and tools are at their peak after 20+ years of evolution. 🤖 You have a 24/7 private tutor - in any topic, any language, any framework (yes, I mean ChatGPT). 💻 Endless paths to explore - frontend, backend, mobile, data, gaming - all accessible if you\u0026rsquo;re curious and motivated. Yes, the landscape is changing - fewer internships, tighter competition. But if you\u0026rsquo;re passionate about building, solving problems, and creating through code, the best time to start is right now. 🚀\n💬 Feeling lost about where to start? Send me a message - I\u0026rsquo;m happy to offer a free consultation to help you get started.\nOriginally posted on LinkedIn.\n","permalink":"/posts/the-software-engineer-career-is-dead/","summary":"The \u0026lsquo;AI killed the dev career\u0026rsquo; headline is wrong. Here\u0026rsquo;s why now is actually a great time to start in software.","title":"The software engineer career is dead"},{"content":"Developers aren\u0026rsquo;t just adapting to AI - we\u0026rsquo;re grieving. Software development has changed forever, and that\u0026rsquo;s a kind of loss. Each of us goes through the 5 stages in our own way 👇\n🩶 Stage 1 - Denial\n\u0026ldquo;It\u0026rsquo;s just another hype cycle.\u0026rdquo; \u0026ldquo;We had LLMs in the 1950s too.\u0026rdquo; \u0026ldquo;My job is way too creative for AI.\u0026rdquo; 🔥 Stage 2 - Anger\n\u0026ldquo;I write better code than this thing.\u0026rdquo; \u0026ldquo;It hallucinates!\u0026rdquo; \u0026ldquo;Stop suggesting console.log()!\u0026rdquo; 🤝 Stage 3 - Bargaining\n\u0026ldquo;Maybe I\u0026rsquo;ll just use it for boilerplate.\u0026rdquo; \u0026ldquo;It\u0026rsquo;s OK as long as it\u0026rsquo;s free.\u0026rdquo; \u0026ldquo;If I supervise it closely, it\u0026rsquo;s basically still my code.\u0026rdquo; 😞 Stage 4 - Depression\n\u0026ldquo;I can\u0026rsquo;t believe I\u0026rsquo;m installing the Copilot extension.\u0026rdquo; \u0026ldquo;What if my PR gets reviewed by a bot next?\u0026rdquo; \u0026ldquo;It writes faster than me — and doesn\u0026rsquo;t complain about Jira.\u0026rdquo; 🧘 Stage 5 - Acceptance\n\u0026ldquo;Let me enable this MCP for my emails.\u0026rdquo; \u0026ldquo;I just asked my AI to write my onboarding doc.\u0026rdquo; \u0026ldquo;Remember when we thought Stack Overflow was cheating?\u0026rdquo; \u0026ldquo;Is there an auto-accept option somewhere in settings?\u0026rdquo; Any of this sound familiar? Where are you in the five stages right now? 👇\nOriginally posted on LinkedIn.\n","permalink":"/posts/ai-coding-tools-and-the-5-stages-of-grief/","summary":"Developers aren\u0026rsquo;t just adapting to AI - we\u0026rsquo;re grieving. A field guide to the five stages, with sample dialogue.","title":"AI coding tools and the 5 stages of grief"},{"content":"Are you into data analytics? 🗃️\nThen if you haven\u0026rsquo;t heard of DuckDB, you\u0026rsquo;re missing out BIG TIME! This bad boy is the newest game-changer in the field. If you haven\u0026rsquo;t checked it out yet - omg, saddle up!\nHere\u0026rsquo;s what I love about it:\n1. Ridiculously easy CSV ingestion and joins\nSELECT * FROM read_csv_auto(\u0026#39;/Users/aston/company.csv\u0026#39;) c JOIN read_csv_auto(\u0026#39;/Users/aston/company_revenue.csv\u0026#39;) cr ON c.\u0026#34;Company Name Column\u0026#34; ILIKE cr.\u0026#34;Some Other Column\u0026#34;; And BAM! You\u0026rsquo;ve got your joined dataset.\n2. Serverless simplicity\nYour database is literally just a file on your machine. Open it in DBeaver and use it like any other DB server — no setup, no daemon, no fuss. You don\u0026rsquo;t even need to install a binary! (But if you want to, it\u0026rsquo;s in basically every package manager.)\n3. Blazing fast\nJoins, aggregations, gigabytes of data - no sweat. Forget uploading to a remote PostgreSQL server - DuckDB flies circles around Postgres and SQLite for analytics.\nGive it a go - once you go DuckDB, you never go back.\n(Okay, maybe not as catchy as \u0026ldquo;once you go Mac,\u0026rdquo; but you get the idea 😎)\nOriginally posted on LinkedIn.\n","permalink":"/posts/once-you-go-duckdb/","summary":"DuckDB is the newest game-changer in data analytics: serverless, just-a-file, and faster than Postgres or SQLite for the analytics queries you actually want to run.","title":"Once you go DuckDB"},{"content":"What do I want? 🤔\nIt\u0026rsquo;s a question I\u0026rsquo;ve asked myself countless times - and one that others have asked me just as often.\nFor years, my answer was confident: \u0026ldquo;Of course I know what I want.\u0026rdquo; Then one day, it shifted to: \u0026ldquo;Wait… do I actually know?\u0026rdquo;\nOver the past 20 years, I\u0026rsquo;ve tried different ways to find that answer. Here are 7 strategies that helped me figure it out - maybe they\u0026rsquo;ll help you too.\n1. Career Break I took two career breaks - each one deeply transformative in its own way. During my second break, I tried streaming PUBG, one of the fastest-growing games at the time. It didn\u0026rsquo;t take long to realize: being on camera wasn\u0026rsquo;t for me. But that clarity was gold.\n2. Travel and Have Fun In 2014, I packed up my life into storage, bought a one-way ticket to Asia, and spent three months exploring. No plans, no pressure - just doing the things I\u0026rsquo;d always dreamed of. It was freedom, pure and simple.\n3. Changing Companies \u0026amp; Roles Fifteen years ago, I dreamed of becoming a manager — it looked cool, important, and like the next step beyond being \u0026ldquo;just the tech guy.\u0026rdquo; The path turned out to be the hardest one I\u0026rsquo;ve ever taken, full of challenges I never saw coming.\n4. Study In 2012, I was bored with my job and felt drawn to psychology. Instead of overthinking it, I went back to school for a second master\u0026rsquo;s degree - and ended up becoming a published scientist.\n5. Personal Projects While my friends were out partying, I often spent nights and weekends building passion projects. Those projects became my playground for learning - about networking, business, and the startup mindset.\n6. Books \u0026amp; Self-Education There\u0026rsquo;s a book for almost everything in life. Over the years, learning from the best through books has been one of my biggest advantages - it\u0026rsquo;s like having mentors on every shelf.\n7. Counselling \u0026amp; Psychotherapy I\u0026rsquo;ve spent six years in therapy and therapy training - facing my inner demons and learning to understand myself. That inner work gave me the courage to step out of my comfort zone and try all six strategies before this one.\nOriginally posted on LinkedIn.\n","permalink":"/posts/7-strategies-to-figure-out-what-you-want/","summary":"Twenty years of trying things - career breaks, travel, study, side projects, therapy - and what each one taught me about the question we all keep asking ourselves.","title":"7 strategies to figure out what you want"},{"content":"🚀 When you need to make small changes to existing code, here\u0026rsquo;s a pro tip 👇\n❌ Don\u0026rsquo;t do this: Try to describe all the changes in a long prompt - it\u0026rsquo;s error-prone and the AI may misunderstand the context. You\u0026rsquo;ll end up trying and failing multiple times. 😩\n✅ Do this instead: Go through your code and add clear TODO comments, like:\n// TODO: Add request validation to this endpoint // TODO: Make this method take an extra parameter // TODO: Refactor this to use async APIs Then use a prompt like:\nGo through the code and collect all TODO items, then for each item implement the changes and update the tests.\n🎯 Why it works:\nThe AI has extra context right in the code You get more predictable and accurate changes Try it out - it\u0026rsquo;s a simple shift that makes a huge difference in codegen workflows. 💪\n👇 Let me know your experiences or tricks in the comments!\nOriginally posted on LinkedIn.\n","permalink":"/posts/ai-code-generation-tips-1-use-comments/","summary":"When you need to make small changes to existing code, don\u0026rsquo;t try to describe everything in one long prompt. Add TODO comments where the changes belong, and let the AI sweep through them.","title":"AI code generation tips #1: use comments"},{"content":"If you ever find yourself reading this page - which you definitely aren\u0026rsquo;t supposed to, because who reads the very first post of a blog anyway? - please send me an email with just the words hello, world in the body, and I\u0026rsquo;ll invite you for a coffee somewhere convenient. We can chat.\nAnd thanks for reading. I really appreciate it.\nFast forward ten or twenty years from now, I hope to find this post and let it fill me with the warm glow of nostalgia - reminiscing about all the great times I had on this site, writing these articles.\nBut to write something of actual value: I plan to cover a lot of ground here, mostly tech, with a focus on the human side. And AI. Or whatever the latest trend happens to be.\nHopefully not jQuery and Internet Explorer bugs anymore.\nOnwards!\n","permalink":"/posts/hello-world/","summary":"First post - a quiet hello and a note on what I plan to write about here.","title":"Echo 'Hello, world!'"}]