Paper 'BlendSQL: A Scalable Dialect for Unifying Hybrid Question Answering in Relational Algebra' accepted to 2024 ACL findings
Research Experience
Started a new role as Principal Data Scientist @ Capital One in October 2024
Background
Data Scientist working in NLP. Currently very interested in methods combining symbolic reasoning with LLMs, constrained decoding, and database systems.
Miscellany
Projects include BlendSQL (Query language for blending SQL and LLMs across structured + unstructured data, with type constraints), BillLit (A React web app to help you split the bill at restaurants), Grammar Guide (Speculative grammar backtracking for LLM decoding conforming to some Lark context-free grammar (CFG))