Full-Stack Engineer - Enterprise AI Agents

A small seed-stage startup in San Francisco builds the infrastructure that keeps enterprise AI agents improving after launch. Their control plane manages the knowledge behind production chatbots and agents and connects how those agents perform in the real world with the knowledge they depend on. For a team this size, the customer list is unusual: they already run live deployments with household-name enterprises, mostly support and IT teams in heavily regulated sectors.

What you’ll own

• The backend of a platform that evaluates agents and keeps improving them
• Data modelling for everything the platform tracks: test sets, experiments, agent traces, knowledge versions and results
• Enterprise integrations, fully yours: auth, webhooks, syncing and backfills, and making them robust and observable
• Features through the whole stack, from design to deployment, monitoring and customer feedback
• Real say in architecture and product decisions, and time with enterprise customers when it matters

• 3-6 years building and running production software
• Backend depth, plus the ability to ship across the full stack
• TypeScript on Node.js or Bun in production, and you’re strong in Postgres
• Experience with Redis, queues, workers or workflow engines
• You’ve built APIs, integrations and data flows for multi-tenant software
• Observability and distributed systems; you enjoy debugging workflows that span many services

Bonus points

• Enterprise integrations you owned end to end
• Exposure to LLM observability, agent evaluation or ML infrastructure
• Startup experience after time at a strong engineering company