I am a data scientist and Python developer with experience in applied statistics, experimentation, and full-stack analytical tooling. My background includes academic research as well as independent software and data projects.
Long-term project launched in 2025. Idea is to crowd-source the generation of a massive (anonymous) mental health dataset that will be useful someday for precision mental healthcare. 1570 participants to date - goal is 20,000 by 2030.
Playlist generation from free-text requests and/or manual fine-tuning. Create custom playlists via youtube API (up to ~200 songs per day until I register with google)
Public facing RAG app for uploading pdfs (research articles) and asking free-text quesitons of them. Modifiable parameters to adjust prompting techique and retreival strategy/scope.
A chatbot specifically prompted NOT to be overly sychophantic or engaging. Bbased on lots of experience with LLMs, and on the ideas of Tristan Harris.