Product Manager, Cloud Agents at Cursor in San Francisco, CA

Our mission is to automate coding. The first step in our journey is to build the best tool for professional programmers, using a combination of inventive research, design, and engineering. Our organization is very flat, and our team is small and talent dense. We particularly like people who are truth-seeking, passionate, and creative. We enjoy spirited debate, crazy ideas, and shipping code. About the role Cursor is entering the third era of AI software development. Developers no longer write most of their code—they direct fleets of agents that do. More than a third of the PRs we merge internally are created by agents running autonomously in cloud VMs. As a Product Manager for Cloud Agents, you will own the product that makes this possible. Cloud agents run on their own virtual machines, take on tasks over hours, iterate and test independently, and return with something quickly reviewable: logs, video recordings, and live previews rather than diffs. Your job is to make this loop fast, trustworthy, and worth relying on for real work. This means owning how agents are provisioned, orchestrated, and monitored. It means designing the experience that lets a developer hand off a task and move on to the next one. It means making the output legible enough that reviewing an agent's work takes minutes, not an hour of reconstructing what happened. EXAMPLE PROJECTS INCLUDE... - Designing the task handoff experience end-to-end: how a developer describes what they want, how much context the agent receives, how the system handles ambiguity, and how the developer moves on without worrying about it. - Owning the agent orchestration model: sandboxing, parallelism, retries, error recovery, and resource allocation. Cloud agents compete for compute—you'll decide how that works. - Building the artifact and review layer: agents return logs, screenshots, video recordings, and live previews. You'll define what artifacts agents produce, how they're surfaced, and how developers verify output without reconstructing each session from scratch. - Shaping the self-hosted and enterprise deployment model: some customers need agents running entirely inside their network. You'll work with security and solutions to design deployment models that meet those requirements without fragmenting the product. - Instrumenting success: task completion rate, time-to-value, cost-per-task, developer trust. Building the measurement layer so the team knows where to invest. - Defining how cloud and local agents hand off to each other: developers move sessions between cloud and local depending on what they need. You'll own the UX for that transition. YOU MAY BE A FIT IF - You have shipped developer tools, infrastructure products, or AI-powered applications as a product manager. - You understand distributed systems well enough to have real conversations with engineers about orchestration, sandboxing, and reliability. - You have strong product intuition for developer experience. You know when something feels right and when it doesn't. - You are deeply technical. You don't need to write production code, but you can read a PR, reason about an architecture diagram, and form opinions about tradeoffs. - You are comfortable making decisions with incomplete information and iterating fast. - You communicate clearly—written and verbal—and can align a team around a direction without a 40-page PRD. - You use AI coding tools regularly and have strong opinions about what's working and what isn't. #LI-DNI