Director of Product Management, AI Agents and Ecosystem

An overview of this role

As the Director of Product, AI Agents and Ecosystem, you'll shape how the GitLab Duo Agent Platform connects to the fast-moving AI agent landscape. Your focus is two fold: you'll define an open, interoperable agent experience that works alongside leading models and agent platforms (including Anthropic Claude, Google Gemini, and OpenAI), and you'll drive adoption of GitLab's agent builder, AI catalog, and out-of-the-box agents and flows. You'll report to the VP of Product, AI and partner closely with engineering leadership and the security and DevSecOps product teams. In your first year, you'll set a clear ecosystem strategy, define how we approach interoperability and multi-agent systems (including tool integrations and Model Context Protocol (MCP) patterns), and create an operating model that helps teams across GitLab build, ship, and iterate on agents customers want to use.

What you'll do

• Lead the strategy for GitLab's agent builder and broader agent ecosystem within the GitLab Duo Agent Platform, with a focus on interoperability and an extensible platform approach.
• Define how GitLab works alongside external model and agent platforms (including OpenAI, Anthropic Claude, and Google Gemini) while creating clear product value for customers.
• Drive the roadmap for agent interoperability, including multi-agent patterns, tool integration, and Model Context Protocol (MCP) considerations.
• Partner with engineering leadership to align on vision, operating model, and execution across distributed teams.
• Collaborate with product teams across DevSecOps and Security to enable domain experts to build high-value agents and flows, and bring them into the Duo Agent Platform.
• Guide the approach for growing adoption and usage of GitLab's agent experiences, including out-of-the-box agents, the AI catalog, and custom and external agents.
• Work closely with partnerships and strategy and operations to shape ecosystem relationships and integration priorities.
• Communicate strategy, tradeoffs, and decisions clearly in an async-first environment, keeping stakeholders aligned from early discovery through delivery.

What you'll bring

• Experience building or leading agent builder products, including defining agent instructions, tool use, and developer experience for low-code, no-code, or code-first workflows.
• Strong understanding of the AI ecosystem, including interoperability patterns across model and agent providers (for example: OpenAI, Anthropic Claude, and Gemini).
• Knowledge of multi-agent systems and approaches for agents from different ecosystems to work together.
• Familiarity with tools and standards for agent interoperability (for example: Model Context Protocol (MCP)) and an open-platform mindset.
• Experience driving adoption of new AI products, including out-of-the-box agents, an AI catalog, and support for external and custom agents.
• Strong cross-functional collaboration skills, including partnering with engineering leadership, partnerships teams, and product teams across DevSecOps and Security to scale agent creation beyond a single team.
• Clear, direct communication and comfort working asynchronously with team members across regions; transferable experience is welcome if you have relevant agentic AI and platform background.

About the team

The Duo Agent Platform team builds the foundations that let GitLab customers use AI agents throughout the DevSecOps lifecycle. We build shared services and integration points that help teams across GitLab create, ship, and run agents and flows, including Agent Builder and ecosystem integrations. We collaborate asynchronously with Product, Engineering, Security, DevSecOps, and Partnerships across regions, with a focus on clear documentation and decision-making. The core opportunity is to deliver an interoperable platform that provides clear standalone value (not just a thin layer over other tools) and an operating model where domain teams can build high-value agents that roll up into the Duo Agent Platform. For more on how we work, see Team Handbook Page.