Lotus Health
Research Engineer
San Francisco · On-site · Full-time · Posted Nov 24, 2025
About the job
Design, evaluate, and deploy AI reasoning and agent systems that turn medical data into accurate, cited, and actionable guidance for patients and clinicians.
What you'll do
- Build agentic workflows and fine-tune models for medical-data tasks, with safety and uncertainty handling.
- Curate and label clinical datasets; own evaluation and safety systems for medical correctness, bias, hallucinations, and performance.
- Bring research into production, optimizing inference and collaborating on retrieval, storage, and data lineage.
What they ask for
- Strong Python and deep experience with PyTorch or JAX.
- Track record of shipping applied ML or LLM features to real users.
- Hands-on experience with RAG, prompt and tool design, evaluation harnesses, and error analysis.
- Solid SQL and comfort with vector stores and retrieval patterns; willingness to work on-site in San Francisco.
Nice to have
- Experience with clinical ontologies and standards such as SNOMED CT, ICD, LOINC, RxNorm, FHIR, and NCPDP.
- Background in RLHF or RLAIF, distillation, program-of-thought, structured tool use, or medical speech and multimodal pipelines.
Tools
Python, PyTorch, JAX, SQL, vLLM, TensorRT, Docker, ECS
AI focus
Applied medical AI research and engineering: agentic workflows, model fine-tuning and distillation, retrieval and tool use, clinical evaluation, safety, and production deployment.
Model-written summary · the original posting is authoritative