About

Adrian Sieradzki

Creator of RLMatrix. Independent developer working on reinforcement learning, engineering tooling, and the gap between research and production.

Hello

I built RLMatrix because I needed reinforcement learning that actually slotted into real engineering systems — not a research notebook. I’m a PhD student at ETH Zurich in the deMello group working on microfluidics, and the tools I needed didn’t exist in the .NET world, so I wrote them. The manifesto is the longer version of why.

Outside of RLMatrix, I also build Kakuso — a Japanese dictionary and handwriting learning tool. More about both projects, plus longer-form writing, lives on sieradzki.io.

If you’re using RLMatrix, extending it, or hitting something that doesn’t work the way you expected, reach out. I read everything.

Find me

Blog — sieradzki.io Long-form writing on RL, engineering, and whatever I'm currently obsessed with.

GitHub — @asieradzk RLMatrix and other open-source work.

X / Twitter — @adisiera Shorter updates, threads, and RL-adjacent commentary.

Email — [email protected] Direct contact for collaborations, consulting, or technical questions.

Working together

RLMatrix is MIT-licensed and free to use in any project, commercial or otherwise. If you’d like help designing an RL system, integrating RLMatrix into a production stack, or sponsoring specific features or algorithms, email me at [email protected].