Ridgefield, CT26 days ago
Work with enterprise RAG and vector search, including designing hybrid retrieval (BM25 + embeddings), chunking and indexing strategies, vector DB tuning (Chroma/FAISS/OpenSearch), document parsing, citation grounding, and latency/quality trade-offs for large internal corpora. Work with legacy-to-python workflow migration with statistical parity, including translating SPSS/legacy analytics into reproducible Python pipelines, creating validation suites to ensure statistical equivalence, and instituting QA/monitoring for ongoing correctness.