Welcome to The Reasoning Trace!

I’m Jonatan Borkowski, a software engineer with over 10 years of experience. After spending the past year diving deep into AI agents and LLMs, I’ve decided to start sharing what I’ve learned.

What to Expect

This blog will cover:

AI Agents

Building reliable, production-ready agents is harder than it looks. I’ll share patterns, pitfalls, and practical approaches I’ve discovered.

LLMs in Practice

Not just theory—how to actually use large language models effectively in real applications.

Software Engineering

Lessons from a decade of building software, now applied to the AI era.

Why Now?

The AI landscape is evolving rapidly. Every week brings new models, frameworks, and possibilities. I want to document my journey and hopefully help others navigate this space.

“The best way to learn is to teach.”

I believe that by writing about these topics, I’ll understand them more deeply—and maybe help you along the way.

A Note on the Name

“The Reasoning Trace” refers to the visible chain of thought that AI models use to reach conclusions. It’s a fitting metaphor for what I want to do here: show my work, step by step.

Let’s explore together.


Have questions or suggestions? Reach out on X or LinkedIn.