Agent Engineer

About the Role

You'll build and own the core agent infrastructure powering Doctours' AI systems—the orchestration layer, tool integrations, retrieval, and evals that run our patient-facing and internal agents. This is a systems and infrastructure role focused on designing agentic workflows that automate clinic-side and internal operations end-to-end, distinct from live conversation AI work. You'll ship reliable agent systems and continuously improve them through monitoring and evaluation.

What You'll Do

• Design and implement the orchestration layer for multi-step agentic workflows across clinic operations and internal processes.
• Integrate external tools and APIs into agent systems, ensuring reliable tool calling and error handling.
• Build retrieval and evaluation infrastructure to measure and improve agent reliability and accuracy.
• Automate recurring manual work by identifying opportunities and shipping systems to eliminate them.
• Monitor agent performance in production and iterate on system design based on real-world behavior.

What We're Looking For

• Strong math and CS fundamentals, demonstrated through elite academic performance, competition experience, or fast self-directed learning.
• Personal projects, research, or open-source work building agent or LLM systems—orchestration, tool integrations, RAG, or evals all count as strong evidence.
• An automation instinct: you spot recurring manual work and naturally think in systems.
• No professional experience required; recent graduates and self-taught builders are welcome.

Compensation & Benefits

$150, 000–$200, 000 base salary + meaningful equity.

Location

Full-time, in-person at our San Francisco headquarters, five days per week.