Projects
Projects and open-source work in AI systems performance: RAG pipeline optimization, vector database benchmarks, and embedding compression for production ML systems.
Collaborations
Working with Ethan Davis at SAS Institute on embedding compression techniques for production ML systems.
Open Source
- LangChain - RAG pipeline optimizations
- Vector DB Benchmarks - Performance analysis tools
- Fine-tuning Utils - Memory-efficient training utilities
Explorations
Interactive demos live with the rest of the research output, including the Linguistic Semantic Chunking explorer.