Staff Product Manager, AI

The role

This role leads product strategy for the agent layer of a new workspace that unifies a company's code and marketing tools. It covers the agents themselves, the infrastructure they run on, external access paths, and the shared context they draw from. The PM serves both developers extending the system and marketers working with it day to day.

What you would do

• Set vision and roadmap for the agent framework, sessions, shared memory, and orchestration core, then drive delivery with engineering, design, and data partners
• Make agents trustworthy by defining disclosure of capabilities and origin, least-privilege permissions, reviewable change output, and gradual expansion of autonomy
• Establish the quality standard for non-deterministic experiences, including evaluation infrastructure, test harnesses, guardrails, and success metrics
• Build shared primitives that multiple Source workstreams can reuse, and sequence work so efforts compound instead of duplicating
• Extend external access through MCP and CLI so outside agents and scripts match embedded agent capabilities and governance
• Shape how persistent, governed project instructions evolve so agents share brand, structure, and business context
• Represent the agent area to leadership and sister teams, resolving cross-functional disagreements without waiting for direction
• Prototype, ship code changes when useful, and keep close enough to implementation to make sound product decisions

What they asked for

• Eight or more years of product management, including ownership of strategy for large, loosely defined product areas
• Deep experience with AI-driven systems and a clear sense of where LLM tools perform well versus where they fail
• Track record of building platforms and APIs that other product teams depend on
• Demonstrated ability to define evaluation, guardrail, and metric frameworks for products with variable outputs and to hold teams to them
• Experience designing both developer-facing tooling and accessible products for non-technical users
• Comfort with hands-on technical work, from reviewing code to running analyses and writing specifications
• Strong influence-based leadership with senior stakeholders across engineering, design, go-to-market, and executives
• Bachelor's degree or equivalent relevant experience

Nice to have

• Hands-on familiarity with code generation tools or programmable interfaces
• Ongoing fluency in emerging AI techniques and how to apply them to team productivity