About the role
You'll work on models and pipelines that draft clinical documentation from real hospital inputs. Accuracy, hallucination control, and eval rigor matter more than demo flashiness.
What you’ll do
- Train, evaluate, and ship NLP/LLM workflows for clinical document generation
- Build evaluation sets with clinicians for specialty-specific quality
- Reduce rework signals (missing sections, inconsistent coding, unsafe phrasing)
- Collaborate with product and engineering on latency, cost, and fallbacks
- Document model behavior, limits, and monitoring for hospital deployments
What we’re looking for
- 3+ years in applied ML/NLP with production systems
- Hands-on with LLMs, fine-tuning, RAG, or structured generation
- Python fluency and solid software engineering habits
- Ability to design metrics that match clinical review outcomes
- Care for privacy and data-handling constraints
Nice to have
- Clinical or biomedical NLP experience
- ICD-10 / discharge documentation familiarity
- Experience evaluating generative models in regulated settings