Senior Data Scientist, Agentic AI Systems

Responsibilities

• Build agentic AI systems for rare disease research workflows, including conversation logic, information-gathering rules, confirmation steps, and downstream handoffs.
• Model outputs with Pydantic using structured output and tool calling so generated fields are typed, validated, and traceable.
• Write, version, and regression-test prompts for clinical and scientific reasoning tasks.
• Build evaluations for open-ended tasks using golden sets, offline regression suites, and model-based graders.
• Keep multi-step LLM workflows responsive under load through asynchronous design, concurrency limits, streaming, timeouts, and failure handling.
• Log request identifiers, latency, errors, and the rationale for automated decisions for later review.
• Translate researcher, clinician, and patient-community needs into data models and system behavior.
• Contribute to manuscripts, conference abstracts, and posters with NIH investigators.

Requirements

• Bachelor’s degree in Data Science, Computer Science, Bioinformatics, Biomedical Informatics, or a related field; equivalent professional experience may substitute. An advanced degree is preferred.
• At least 5 years building and operating production software or data systems, including at least 2 years shipping dependable LLM-powered applications.
• Experience with structured output and tool or function calling.
• Experience evaluating systems without a single correct answer using golden sets, offline regression suites, or model-based graders.
• Ability to own a service end to end, from schema design through deployment and operation.
• Ability to obtain and maintain a Public Trust Security clearance.
• Python with FastAPI, Pydantic, and pytest.
• LLM application engineering, including provider APIs and gateways, prompt and context design, structured generation, tool use, and tracing.
• PostgreSQL, including embeddings or vector search alongside relational data.
• Asynchronous and concurrent Python and streaming results to clients.
• Containers and Kubernetes.
• Git-based collaboration and CI/CD.

Nice-to-haves

• Typed agent frameworks such as Pydantic AI, LangGraph, or OpenAI/Anthropic agent SDKs; MCP.
• LLM tracing and evaluation tools such as Langfuse, LangSmith, Arize Phoenix, or Braintrust.
• Production serving of open-weight models with Ollama or vLLM behind LiteLLM.
• Biomedical ontologies and controlled vocabularies, including MONDO, HPO, UMLS, MeSH, and OBO Foundry resources.
• Rare disease, clinical genetics, or translational research experience.
• Experience working with clinicians, curators, or patient advocacy organizations.
• Published or presented engineering work, including papers, conference talks, preprints, technical blog posts, or open-source contributions.
• Prior or current NIH experience.

Compensation and Benefits

• Salary: $130,000–$150,000 annually.
• 100% employee medical, dental, and vision coverage.
• Paid time off and paid holidays.
• 401(k) match up to 5%.
• Educational benefits, employee referral bonus, and flexible spending accounts.