maximalist generalist building ml systems that work. focused on practical llm applications. llm-maxxing as a lifestyle choice.
- ml engineer
- architecting production-ready ai solutions
- working on making large models actually accessible
- local-rag - retrieval-augmented generation system
- groqLive-release - low-latency voice transcription/response pipeline
- image-comparator - performance-focused image processing tools on a static webpage, client-side javascript
my latest blog post: AI stands for artificial illiteracy
open to collaborating on interesting ml projects, especially if you're into pushing the boundaries of what's possible with llms


