I'm a software engineer from France ;) I've studied maths/physics, and now I enjoy programming a lot!
My interests are quite diverse, spanning traditional software engineering, modern machine learning and nlp, but also lower-level intricacies.
Over the past 2 years, I've worked on several projects:
- an HTTP/1.1 server, a (forwarding-)DNS server
- a tree-walk and a VM-based interpreter
- a (boolean) keyword search engine
- a neural net lib from 'scratch' (one of the things I did while at RC)
I also started writting my personal website where you can read my first poasts.
I've been using Rust for nearly all my side projects and enjoying this new language a lot. However I'm using Python daily at work and love this language as well.
Recently, I was able to contribute to some of my favourite open-source projects. In particular, I'm interested in projects with Rust / Python interop.
-
Pydantic: I improved error messages for validation and serialization (Rust / Python interop).
-
Pydantic AI: I wrote an issue and then implemented a solution myself. I iterated through multiple code reviews to get it right, writing new tests and a new documentation section. I had to think carefully about the new feature public API and type design. (Python)
-
Instructor: I wrote an issue and then a patch solving a sad JSON parsing bug I ran into at work. Typical case of LLMs being very confident and very wrong at the same time. (Python)