Engineering Manager, Context (agentic Search)

Notion Notion · Enterprise · San Francisco, CA · Engineering

Engineering Manager to lead the Context team, responsible for systems and product primitives enabling user and agent access to relevant context within Notion. The role involves managing a team focused on both the reliability of current search/context systems and the development of new agentic capabilities, including tool use and iterative retrieval. The position requires balancing foundational platform work with rapid iteration on agentic features, and partnering across AI, product, and infra teams.

What you'd actually do

  1. Build and lead a high-performing team responsible for context, memories, and ranking that power agent workflows across Notion.
  2. Drive a roadmap that improves the end-to-end reliability of agents acquiring context across Notion + connectors.
  3. Raise the bar on context quality through evaluation + experimentation, and ship improvements that move core user metrics.
  4. Balance foundational platform work (latency, cost, stability) with rapid iteration on new agentic capabilities.
  5. Partner deeply across AI, product, infra, and partner ecosystems to unblock delivery and set clear ownership boundaries.

Skills

Required

  • Experience leading engineering teams working on search/retrieval/relevance, AI product, or adjacent platform problems.
  • Experience building agentic or tool-using systems
  • Strong product + systems judgment in ambiguous environments; can make crisp tradeoffs and keep execution moving.
  • Comfort owning a roadmap that requires heavy cross-team coordination and stakeholder alignment.
  • Sufficient technical depth in modern retrieval (lexical + semantic, embeddings, reranking) and where LLMs fit (query generation, tool orchestration, evidence selection).
  • Track record building collaborative, supportive, and high-expectation team cultures.

Nice to have

  • Experience with multi-source retrieval and permissioned enterprise data.
  • Experience establishing eval frameworks for AI/agent outcomes.
  • Experience with classic information retrieval metrics and machine learning techniques.
  • Has led teams through rapid scope changes and evolving org boundaries.

What the JD emphasized

  • production-grade search product
  • agent layer
  • tool use
  • evaluation frameworks

Other signals

  • AI-enabled search traffic is growing
  • Agent layer accounts for the majority of AI-enabled search traffic
  • Drive a roadmap that improves the end-to-end reliability of agents acquiring context across Notion + connectors
  • Raise the bar on context quality through evaluation + experimentation
  • Balance foundational platform work (latency, cost, stability) with rapid iteration on new agentic capabilities
  • Agentic retrieval approaches that treat “finding” as a sequence of actions (tool use, query generation, iterative retrieval), not a single ranked list
  • A durable measurement loop: evaluation frameworks, instrumentation, and experimentation practices to improve functional correctness and user-perceived usefulness