Product Engineer, AI (Member of Technical Staff)

About the role

3Y Health builds an AI-native operating system for healthcare practices, serving ABA therapy practices now and expanding into speech, occupational and physical therapy and psychiatry. As a Member of Technical Staff (product engineer, AI), you own features end to end: you work with clinicians and operators to scope the problem, design and build the solution across the stack, deploy it and follow its performance in production. The engineering team is flat, with no sub-teams, and the company uses AI coding agents daily.

Key responsibilities

• Design and build AI agents and automated workflows for clinic operations such as insurance benefits verification, prior authorizations, patient intake, claims and clinical documentation.
• Build features for clinicians, patients and families, and for internal operations, on web and mobile.
• Turn ambiguous problems from 0 to 1, defining the problem, writing the spec and shipping a first version.
• Design services, data models and APIs and manage infrastructure as code.
• Own services in production: deployments, monitoring and incident response.
• Write evals for model-facing features and track workflow metrics.
• Handle protected health information safely, with least-privilege access, no PHI in logs and auditable changes.
• Maintain the skills, plugins, MCP integrations and review bots used by engineers with Claude Code and Codex.

Required profile

• 5+ years of experience building and shipping production software.
• Track record of owning products or features end to end.
• Hands-on experience with AI harnesses (project instructions, tools, MCP servers, skills, hooks, subagents, evals), not just model API calls; able to share what was set up or built.
• At least one personal project published online and usable by anyone, ideally a SaaS with payments.
• Good product sense, clear written communication and high agency.

Required skills

• Strong full-stack Python and TypeScript, across frontend, backend, databases and cloud.
• Comfortable in a large, established codebase.
• Hands-on experience building with LLMs: tool use, agent orchestration, structured outputs, retrieval and evals (LangGraph or similar).
• Daily use of AI coding agents such as Claude Code, Codex or Cursor.
• Nice to have: healthcare or regulated industry experience (HIPAA, payer and claims workflows, EHR integrations, HL7/FHIR), Temporal, React Native, running LLM evals in production, and open-source or internal tools.