Senior Machine Learning / AI Engineer (with commercial experience)

Senior AI/LLM engineer — RAG, agents, and eval. Owns the entire AI lane. Open source and frontier models. This is the hard-dependency lane — if it slips, the product's core promise (cited, accurate-by-construction memos) doesn't ship. Must have: production RAG with reranking, agent orchestration (hand-rolled supervisor patterns, not just framework glue), and a real evaluation discipline (golden sets, MRR/recall, LLM-as-judge). Strong Python + TypeScript; comfortable with provenance/citation-at-write-time and "emit missing, never guess" correctness thinking. Anthropic/Claude SDK and prompt-caching familiarity is a plus. Not an ML researcher or a model-trainer — this is applied retrieval/agent/eval engineering on a deadline. Bias toward someone who ships and can prove accuracy, not someone who fine-tunes. Signal to screen for: shipped a cited/grounded RAG system, can explain how they'd stop a model from inventing a number, and treats evaluation as a first-class deliverable.