Qualifications
What You’ll Do:Lead initiatives to enhance Datadog’s public-facing MCP server, enabling intelligent agents to seamlessly discover and interact with our services.Design and implement agentic tool surfaces suitable for both evaluation and production across diverse AI agents.Develop and maintain sophisticated evaluation pipelines to assess agent performance on Datadog workflows, including investigations, incident triage, and metric queries.Analyze and resolve failure cases by examining tool output, improving query parsing, and refining agent feedback mechanisms.Collaborate with Applied AI and internal teams to establish unified standards for tool integration and data access.Who You Are:A seasoned Staff-level engineer with a robust background in applied AI and agentic programming, including experience with LLM-powered automation pipelines and orchestration frameworks (e.g., LangChain, LangGraph, CrewAI).Comfortable navigating ambiguous and rapidly changing environments, with the ability to autonomously define and prioritize tasks.Skilled in developing evaluation frameworks for LLM agents or AI systems, encompassing metrics design and data instrumentation.Bonus: Experience with the MCP standard or contributing to agent-compatible tools, familiarity with constructing and evaluating ReAct agentic loops.Possesses strong systems thinking, capable of reasoning across multiple agents, tools, and user scenarios.Passionate about advancing agent-augmented software and eager to influence the development of evolving interfaces.
About the job
The MCP team at Datadog is expanding its capabilities by enhancing our MCP Services, a platform that facilitates interactions between external agents such as Claude and Cursor, as well as our internal Datadog AI Agents, and Datadog's data. As a Staff Engineer, you'll be instrumental in refining this agent-friendly interface, fostering agent-driven workflows across various data types including metrics, logs, and incidents. Your contributions will be pivotal in developing advanced evaluation frameworks for agentic systems and shaping the future of agent-tool interactions. This is a dynamic field with significant impact, perfect for someone enthusiastic about influencing Datadog's applied AI trajectory.
At Datadog, we cherish our office culture, which fosters collaboration, creativity, and strong relationships. We embrace a hybrid work model, allowing our employees to achieve a work-life balance that suits their needs.
About Datadog
Datadog is a leading monitoring and analytics platform for developers and IT operations, providing observability across cloud applications and infrastructure. Our team is dedicated to innovation and collaboration, creating a vibrant workplace culture that values diverse perspectives.