Hi!

I'm Sofie.

I am a Software Engineer who just loves coding, debugging, documenting, testing, nitpicking and fine-tuning. I prefer code to be sufficiently robust and modular, and I care a great deal about user experience (UX) and quality assurance (QA). I always love to do the simple thing first 🛖, then build up the complexity as required 🏛️. I have a PhD in Natural Language Processing (NLP) 🎓 and I've been working in AI ever since my Master's back in 2007 - though we just called it Machine Learning back then ;-)

I'm passionate about open-source 🐍

Some of the OS projects I've helped maintain include:

• Andrej Karpathy's autoresearch to let AI Agents run research autonomously (March 2026 - now)
• Andrej Karpathy's nanochat: an educational repo implementing GPT2 capabilities from scratch (Oct 2025 - now)
• FastAPI and friends by Tiangolo, including FastAPI to build APIs, Typer to build CLIs, SQLModel to interact with SQL databases from Python, and several others (2024 - now)
• Some personal vibe-coded hobby apps, including a fun language learning tool for my Arabic lessons, and a simple calorie tracker. (2025 - now)
• Explosion's NLP framework spaCy & spacy-transformers & spacy-llm, and deep learning library Thinc (2019 - 2023)
• Huggingface's neuralcoref, a Coreference Resolution system they built before turning their attention elsewhere... (2019 - 2021)
• Created & published now entirely outdated open-source software back at uni (2007 - 2015)

Other OS tools I know quite well / use regularly:

• pytorch, transformers, tokenizers, datasets, kernels
• wandb, matplotlib, numpy, scipy
• pydantic, rich, click
• sentence-transformers, scispacy, scikit-learn
• ruff, ty, black, mypy, coverage
• uv, pip, pytest, smokeshow, setuptools

If you want to follow my writings about the day-to-day life of an OSS maintainer, checkout The Diff & The Merge substack!