- Architect and continuously improve RAG pipelines by retrieving clinical context and optimizing retrieval layers for clinical relevance.
- Design and build automated evaluation frameworks and datasets to assess AI recommendation quality beyond standard metrics.
- Identify and mitigate gaps in foundation model capabilities through improved retrieval, prompt engineering, and output validation.
- Monitor production AI outputs for quality, drift, and failure modes using built-in evaluation infrastructure.
- Collaborate with clinical leadership and BCBAs to translate clinical domain knowledge into robust technical requirements.
PythonNLPLangChain