Member of Technical Staff - Agent Infrastructure

Probook is the AI operating system for home services. We started with dispatch, the decision that determines how a $750B industry makes its money, and have since built across the rest of the front office. Our customers do $10M to $100M a year in revenue and run their operations on us.

Dispatch was the first step. The harder problems are ahead of us.

Our team: 2x IMO medalists, 3x ICPC World Finalists and coaches, a Columbia professor, former founders, MIT, Waterloo, GT, and engineers from Citadel Securities, Stripe, Palantir, Two Sigma, and Meta.

• Product infrastructure. Build and operate the infrastructure behind our real-time decision systems — durable execution, workflow orchestration, reliability, and performance for products customers run their business on.

• Agent infrastructure. Build the sandboxes, evaluation infrastructure, and harnesses our engineers use to develop and ship agent workflows, including the compute environments agents execute in.

• The developer platform. CI/CD, deployment, local development, and internal tooling — the abstractions that let the team ship in days rather than quarters.

• Reliability and observability. Instrumentation, monitoring, incident response, and debugging tooling for systems whose failure modes are often probabilistic rather than deterministic.

• Cost and scale. Cloud and inference costs as the product footprint grows, and capacity and architecture decisions made before they become bottlenecks.

• Partner across engineering. Work directly with the product engineers building on your infrastructure, and with the customers affected by it.
  

• 3–6 years of engineering experience with meaningful ownership of infrastructure you built rather than inherited.

• Zero-to-one experience. You've built infrastructure from scratch at a startup, or stood up new systems on a team inside a larger company.

• Strong infrastructure fundamentals across distributed systems, cloud infrastructure, networking, CI/CD, and observability.

• Real software engineering. Infrastructure at Probook means writing code and owning complex systems end to end, from design through production.

• Breadth over depth. You'd rather build new systems than operate mature ones, and you move up the stack when that's where the problem is.

• Judgment. You know when a system deserves careful architecture and when it deserves to be finished this week.