SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Patronus AI is a frontier lab developing simulation research and infrastructure to accelerate progress toward human-aligned AGI. The company is behind influential AI evaluation research including FinanceBench, Lynx, SimpleSafetyTests, CopyrightCatcher, and Humanity's Last Exam. Founded by former researchers from Meta AI, Amazon AGI, and Google, the company serves foundation model labs and Fortune 500 enterprises like Adobe, backed by investors including Lightspeed Venture Partners and Notable Capital.
As Senior Product Manager, you will own the product strategy and execution for the internal platform and tools that enable the team to turn frontier-lab demand into delivered RL environments. This includes the hosted environment platform, agentic tooling for building and QA-ing application clones, expert onboarding, delivery tracking, and the spec pipeline feeding engineering.
Key responsibilities include: owning the end-to-end roadmap for internal platform and tools, sequencing against company OKRs so product is never the bottleneck; writing PRDs that engineers trust and build from—versioned, frozen at kickoff, explicit about MVP scope and deliberate cuts; working with internal customers (environment engineers, SME ops, QA, GTM) to identify pain points and translate learnings into a ranked backlog; making usage data the basis for decisions by instrumenting tools and killing features that don't earn their keep; closing the loop between QA and spec by fixing pipeline failures, not just tickets; applying the "if an agent can do it" rule to your own function by building agentic workflows for spec generation and review; ruthlessly scoping to ship the smallest thing that changes an internal customer's week; and staying hands-on by running agents, prototyping with AI, and building small evals to confirm feature impact.
The role operates on a core principle: if an agent can do it, don't assign it to anyone else. Your users are as often AI agents as people, and products should default to agentic workflows with humans as the exception. This is not a backlog-administration role—you should not manage a product you don't use. The product function sits on the critical path for the company's OKR roadmap, and your job is to make specs the thing engineers reach for first: grounded in real usage rather than opinion, with explicit trade-offs and buildability constraints.
