Member of Technical Staff, Agentic Systems

GenMD•Palo Alto, CA

4d•$180,000 - $250,000

About The Position

GenMD is a Stanford spin-off building AI and data infrastructure for healthcare. Its platform makes large longitudinal datasets—including claims, health records, imaging, and clinical notes—usable for researchers and frontier AI while protecting patient privacy. GenMD is revenue-generating and well funded, and operates as a small, fast-moving team. A growing share of GenMD’s product runs on agents, beginning with Mira, an AI research assistant on Healthcare Datahub. Mira helps researchers find datasets, explore cohorts, and write analysis code. The Member of Technical Staff, Agentic Systems will design and ship the systems these agents run on and own them in production, serving real researchers and working with real data. The role works directly with the founders, researchers, and partners.

Requirements

At least 3 years of experience building production backend systems.
Strong proficiency in Python or TypeScript.
Hands-on experience shipping LLM applications or agents, including tool use, structured outputs, and evaluations.
Experience with cloud data infrastructure; GCP and BigQuery are preferred.
Experience building secure, well-tested services.
Ability to own systems end to end and move quickly without lowering the quality bar.

Nice To Haves

Experience with healthcare or other regulated data, data governance, or access control.
Experience building code-generating or data-analysis agents.

Responsibilities

Build and own agent infrastructure, including orchestration, tool calling, retrieval over data catalogs and dictionaries, retries and fallbacks, and human-in-the-loop review.
Ship agents into production and improve them based on real usage, including prompts, tools, guardrails, and cost and latency trade-offs.
Build agent evaluation systems, including offline evaluations, online metrics, and regression detection.
Design services, data models, and APIs that keep agent actions auditable, permissioned, and safe around sensitive healthcare data.
Work with founders, researchers, and partners to determine what agents should do and where human review should remain in the loop.