Full-Stack Engineer, Backend-Leaning (AI Agent Tooling)

A developer-tools startup that helps teams building AI agents understand why those agents fail in production is hiring an engineer comfortable across frontend and backend who leans toward server-side work. You will build product from the interface down to the data layer and spend time directly with customers’ AI teams.

What you’ll build

• Central product features for investigating agent behavior, running evaluations, debugging trajectories, verifying outcomes and replaying runs
• Solutions you define yourself: talk to customers, learn where their agents break, decide what should exist, then build it and keep refining until it holds up
• Server APIs, data modeling, tooling for internal use and polished product screens, with extra weight on the backend
• An in-product AI agent that carries out large-scale investigations over production traces
• Platform foundations: team workspaces, access control, billing and metering, SDKs and command-line-first developer workflows
• Product decisions informed by what production AI teams struggle with

Who thrives here

• Five to eight years of full-stack product engineering, backend-leaning
• Has shipped user-facing product end to end, with both React/Next.js frontend and backend work
• Clean, consistent TypeScript/React plus Python or Node code
• Recent time at a strong product company or a well-funded Series B-plus startup
• Plus: background in AI agents, LLM products, developer tools, observability, or platform features like permissions and billing

On-site in the San Francisco Bay Area. Visa transfers considered; case-by-case review for exceptional candidates.