Senior Data Scientist, Agentic AI Systems
Core
Building conversational agentic AI systems that act as an interface between patients/families and rare disease research infrastructure, capturing validated structured data to accelerate scientific discovery.
Role type
Senior individual contributor (IC) agentic AI systems engineer
Builds
Multi-turn conversational AI workflows with validated structured outputs for rare disease research
Domain
Biomedical informatics / Rare disease research
Deliverable
production ML models
Required skills
Agentic workflow engineering, structured output and tool/function calling, LLM evaluation with golden sets, end-to-end service ownership, asynchronous/concurrent Python, prompt engineering, data modeling
Preferred skills
Typed agent frameworks (Pydantic AI, LangGraph), LLM tracing/evaluation tooling, serving open-weight models, biomedical ontologies (MONDO, HPO, UMLS), rare disease domain knowledge
Technologies
Python, FastAPI, Pydantic, pytest, PostgreSQL, Kubernetes, Git, CI/CD, Langfuse, LangSmith, Arize Phoenix, Ollama, vLLM, LiteLLM
Responsibilities
Design conversation logic and confirmation steps for agentic AI; implement structured output and tool calling with Pydantic; write and test prompts for clinical reasoning; build evaluation frameworks for non-deterministic tasks; ensure responsive multi-step LLM workflows; log system decisions and reasoning; translate researcher/clinician needs into data models; contribute to scientific manuscripts and abstracts
Seniority
Senior, hands-on IC
