On April 13, 2026, I received my AI_Devs 4 Builders certificate. 🎓
You could stop there. A certificate is a certificate — something for the CV, something to click on. But this time I want to describe it differently. Because behind that piece of PDF there is something worth saying out loud.
2018. A Decision. ⚡
Eight years ago I decided to become a programmer.
No CS background. No mentor. I started with a Complete Web Developer course on Udemy — completed on August 24, 2018. Just a strong conviction that I wanted to build things. That code gives freedom — the ability to create something from nothing, using only logic and time.
Today — eight years later — I code and solve real problems: automating processes, building integrations, creating tools for składmuzyczny.pl, where I also work selling guitars in the best music store in Krakow. Yes — you can code and sell guitars at the same time. You can. I have several production projects behind me: Charopis, Melody Bridge, Groove-2026, SupplyScribe and others. These are not "portfolio" projects. They are working tools that solve real problems for a real company.
The AI_Devs 4 certificate is one more brick in that story. I have been part of AI_Devs since the very beginning — since the first edition in 2023. AI_Devs 3 I did not finish due to circumstances beyond my control, but I came back for the fourth edition and this time — the certificate is here. It is also a good moment to thank the course leads — Adam Gospodarczyk, Jakub Mrugalski and Mateusz Chrobok — who created this epic course for the fourth time.
What Does This Certificate Confirm? 📜
AI_Devs 4 Builders is a five-week cohort-based course built around one question: how do you build production AI systems that actually work?
The certificate confirms competencies in four main areas:
- Working with Large Language Models in Code — calling LLMs through APIs, steering behavior via prompts and code, structuring responses using JSON Schema, Function Calling, CLI and MCP tools, multimodality (image, audio, video), security (prompt injection), cost and reliability.
- Context Engineering — the difference between prompt and context engineering, managing context in LLMs, caching, generalized tools for agents, balancing static and dynamic information, context compression, meta-prompts, the non-deterministic nature of LLMs as an advantage.
- Observability and Evals — architecture designed with observability in mind, offline and online evaluations, datasets for prompts and agents, filtering and moderating behavior, improving reliability and performance of the entire system, tooling integrations.
- Building Production Apps — designing, maintaining, and scaling generative app architecture, multi-agent systems running in the background, tools that integrate with existing logic, optimizing business processes with human-in-the-loop, informed tech stack decisions, AI that matches real business requirements.
Sounds like a dry list? It is. But behind each point stands at least one concrete task that had to be built and shipped. Over 25 hands-on projects in five weeks. 🔥
It Was Hard Work. And That Is Fine. 🐂
Honesty is key here: the course is enormous. The material — vast. This is not training you "check off" and move on.
I completed it. But the real work starts now.
That is the nature of AI_Devs-class material — you work under pressure, in sprints, you learn by building, and then you leave with a certificate and a list of things you need to revisit calmly. That is how learning sticks. Not "watched it, got it" — but "built it, broke it, rebuilt it, came out wiser."
The course was also unique in its format — gamification elements where we literally "saved the world" by programming. Sounds like marketing? Maybe. But it worked. The narrative context gave weight to the tasks. Every agent, every system, every tool — was a piece of something larger. We combined the principles of classical programming with autonomous agents powered by LLMs. It is neither "just coding" nor "just AI" — it is a completely new kind of engineering.
What Does This Open Up for Me? 🔭
Let me be honest — I do not know exactly. And that is exactly what is exciting.
A few things I can assess with reasonable confidence:
1. Autonomous agents will transform my projects. Charopis today is a content generation engine — and it is no coincidence that it was born during the first AI_Devs edition, in 2023. Learning by building works. After AI_Devs 4, I see Charopis as an agent system that monitors its own quality, self-corrects, detects when something is off. I do not know yet exactly how or when — but it is interesting. And that is exactly what drives me.
2. MCP is the new integration standard. Model Context Protocol, one of the central topics of the course, is a communication protocol that changes how tools and agents talk to each other. In a year this will be the default way of thinking about integration. I can implement it right now.
3. Observability and evals are the new "unit tests". I learned to think about AI systems not as black boxes, but as structures you need to measure, evaluate, and improve. It changes the approach to the entire software development process.
4. Context Engineering is a new discipline. The most important thing I took from the course: not "how to write a good prompt", but "how to manage information in an AI system so the results are predictable." A subtle but fundamental difference.
Will this translate into concrete projects? Absolutely. When and exactly how — I do not know yet. First I need to go through the material again, calmly, without the sprint pressure. Then comes the implementation. And then we will see. That is the best part — I still do not know what will come out of this. And that is a good answer. 😄
8 Years Ago. Today. Forward. 🚀
I remember the moment in 2018 when I made the decision. I knew it would be hard. That it was a long road. That there would be countless moments when nothing works and you sit there not knowing why.
But I also knew that code gives real possibilities. That every skill acquired stays with you.
Today I look at working projects, at the AI_Devs 4 certificate, at what still lies ahead — and I feel exactly what I felt in 2018. Excitement. Curiosity. Readiness.
The decision made 8 years ago was a good one. And it keeps paying off.
Back to building. 🔨