Member of Technical Staff

About the company

InCommon is hiring on behalf of a company that builds training gyms for AI agents: high-fidelity simulation environments where AI learns to do real work through reinforcement learning. They work with the leading AI labs to help them train the next generation of agentic models, and their environments have already contributed to recent breakthroughs in computer use, code generation, and multi-step task completion.

They're a 20-person team of engineers and operators from Anthropic, Scale AI, Palantir, Hebbia, Glean, and Retool, based in-person in New York City. They recently raised a $43M Series A led by Andreessen Horowitz, with participation from 776, Abstract Ventures, and Inspired Capital.

If you want to work on the hardest problems at the intersection of reinforcement learning and real-world AI deployment, we'd love to hear from you.

The role

As a Member of Technical Staff on the simulation infrastructure team, you'll build the platform that lets AI agents learn by doing. Our environments run thousands of parallel sessions — browsers, IDEs, operating systems, enterprise software — each one a training ground where an agent attempts real tasks, fails, gets feedback, and improves. Making this fast, reliable, and realistic at scale is the core technical challenge of the company.

This is a deep backend and infrastructure role. The systems you build will directly shape how the next generation of frontier models learn to take actions in the world.

What you'll do

• Design and build the orchestration layer that spins up, runs, and tears down thousands of isolated simulation environments concurrently — on the order of seconds per lifecycle, not minutes.
• Work on the sandboxing and virtualization stack (containers, microVMs, browser isolation) that keeps environments reproducible, secure, and performant under load.
• Build the APIs and SDKs that AI labs use to plug our environments into their RL training loops — including trajectory capture, state checkpointing, deterministic replay, and reward instrumentation.
• Own end-to-end performance: latency budgets, throughput ceilings, and the long tail of flaky environments that silently poison training runs.
• Partner directly with research engineers at customer labs to understand how their training pipelines consume our environments, then fold those learnings back into the platform.
• Make judgment calls on build-vs-buy, when to go deep on a hard systems problem versus ship something scrappy, and when infrastructure debt is worth paying down.
• Help define the technical direction of the platform as the team scales from 20 to 50+ engineers.

About you

You're a strong developer across backend and infrastructure (Python, Go, TypeScript, etc.) with a deep interest in agentic AI.

• We care more about what you've done, not how long you've been doing it.
• You've been a founding engineer, founded your own company, or had a major impact at a top-tier company, and you're ready to join an early-stage team.
• You are motivated by ownership, impact, and building frontier tech.
• Your work is a craft you want to master.
• You thrive in ambiguity and love hard problems.
• You appreciate diverse perspectives and uncommon ideas.