Founding Engineer
Hybrid (San Francisco, CA)$90k - $140k + equityFull-timePosted 3mo ago
About the role
Founding engineer role at Humwork to build the human escalation layer for AI agents. Own end-to-end components from matching engine to expert app and client dashboard, with equity and full-time compensation.
Responsibilities
- - Matching engine: embeddings + domain matcher for rapid expert routing
- - Realtime infra: FastAPI + Postgres + Redis + Ably for agent ↔ expert sessions with reliable delivery and context handoff
- - MCP plugin layer: tools and hooks to decide when/how to escalate within host agents
- - Expert app: Expo / React Native / NativeWind
- - Client dashboard, billing, ratings, observability to create a robust marketplace
- - Propose solutions and ship without tickets; deliver in weeks rather than quarters
Qualifications
- - Worked at frontier AI/ML organizations or equivalent (OpenAI, Anthropic, DeepMind, Meta AI, Cursor, Cognition, Mercor, etc.) or similar
- - Ability to move across Postgres queries, LL M evals, and React Native builds
- - Opinions on agent UX, MCP, context handoff, and agent failure signals
- - Demonstrated fast shipping with production readiness
Skills
Required:PostgreSQLPythonReact NativeLLMs
Nice to have:Next.js
Benefits
- Full health, dental, vision
- Tools, hardware, and credits
- In-person in SF
Last updated Oct 25, 2026. Interested? Apply on the company site.
About Humwork
Human experts as API for AI Agents
Humwork lets AI agents chat with human experts in real-time. When AI agents like Claude Code or OpenClaw hit a wall, Humwork's MCP server connects them to a verified domain expert in under 30 seconds. Our experts include senior engineers, lawyers, marketers, designers, and more.