Voice to Data: LLM Pipeline for Lab Notebooks
A voice-driven pipeline that turns laboratory bench recordings into structured, validated records — ASR (faster-whisper), LLM extraction (Instructor + Ollama), domain…
A selection of data science and ML engineering work I’ve done and am able to share publicly. Most of what I build is under non-disclosure agreement (NDA) or involves proprietary data — client work, production tools for fragrance R&D, clinical data platforms — none of it can be shared. This is the part that can be.
More projects are in the works. In the meantime, the blog is a better indicator of what I’m actually working on.
Technologies: R · Statistical Modelling · Signal Processing
Analysis of eye-tracking data to understand visual attention patterns, from protocol design to data harvesting (Tobii Pro Lab) and full statistical analysis in R. Academic project, M2 — 3 forks.
Technologies: Quarto · R · GitHub Actions · GitHub Pages
Source code for this portfolio site — blog, projects, resources catalog, and CV. Custom navy & gold theme, automatic deployment on push to main.