Senior Engineering Manager - Agent Platform, AI Platform

Netflix Netflix · Big Tech · United States · Remote · Data & Insights

Senior Engineering Manager to lead the Agent Platform team at Netflix, responsible for the foundational runtime and execution environment for all agents at Netflix. This includes memory, identity, permissioning, discovery, registry, evaluation, and tool interfaces. The role involves setting technical strategy, hiring and developing engineers, making build/adopt/partner decisions, and turning complex agent problems into platform primitives. The team operates at member scale and focuses on enabling agents to improve over time through feedback loops and evaluation gates.

What you'd actually do

  1. Own the charter and technical direction. Set the platform's technical strategy and roadmap — the agent-specific foundation every agent depends on, and the selective runtimes worth operating directly — and decide where to invest ahead of demand.
  2. Lead and grow a high performing team. Hire, develop, and retain excellent engineers; set a high bar; and create the conditions for a small team to have outsized impact.
  3. Make sharp build, adopt, or partner calls. Decide which capabilities the platform owns and which it draws from partner teams, define the interfaces between them, and keep a small team focused on the highest-leverage layer.
  4. Turn the new hard problems into primitives. The most durable problems in this space are the ones that let agents get better from their own experience: capturing each run's full reasoning-and-tool-call trajectory as structured, replayable data; turning real-world outcomes into evaluation signal; memory that compounds across runs; and a feedback loop that turns that signal into improved behavior, promoted behind eval gates and safe rollout. Make these first-class platform capabilities with real contracts — so agents built on the platform measurably improve over time, not just scale.
  5. Operate as part of the AI Platform leadership team. Set direction alongside peer leaders across the AI Platform and partner orgs, represent the team in senior technical forums, and build the cross-team alignment a foundational platform depends on.

Skills

Required

  • Experience leading and growing engineering teams that build and operate platform or infrastructure at scale
  • managing senior individual contributors and/or managers, with a track record of developing strong engineers
  • Hands-on experience designing, building, and operating agentic systems in production
  • agent runtimes and harnesses, tool use, memory, and evaluation
  • genuine technical depth
  • Breadth across business applications: you've built or led agents spanning a range of use cases, enterprise and internal applications (developer tooling, operations, knowledge work) or consumer-facing applications at scale
  • A strong engineering foundation in distributed systems, cloud, and ML/AI infrastructure
  • deep enough to lead a senior team's hardest architectural decisions and earn its trust
  • A track record of owning a platform's charter
  • setting its direction and holding its scope against pressure to do everything
  • making build-versus-partner decisions on a multi-year cost horizon
  • Cross-organizational credibility: a record of building shared accountability with senior leaders across adjacent platform, infrastructure, and security teams, and navigating ambiguous ownership boundaries
  • Strong written and verbal communication
  • comfort with ambiguity
  • able to lead both 0-to-1 and 1-to-100 work
  • effective across a distributed (US) team

Nice to have

  • Experience with the emerging class of agent concerns: agent identity and permissioning, guardrails and safety, and cost governance for LLM and agent workloads.
  • Experience running a platform that other engineers consume as a product — paved paths, migrations, deprecations, and developer experience.
  • Credibility in a relevant open-source ecosystem (agent frameworks, runtimes, or developer tooling).

What the JD emphasized

  • agent-specific foundation
  • runtime and execution environment
  • memory
  • identity and permissioning
  • discovery and registry
  • evaluation
  • tool/MCP interfaces
  • paved path
  • member scale
  • capture reasoning and tool-call trajectory
  • structured, replayable data
  • real-world outcomes into evaluation signal
  • memory that compounds across runs
  • feedback loop
  • improved behavior
  • eval gates
  • safe rollout
  • first-class platform capabilities
  • measurably improve over time
  • judgment
  • what to own
  • what to hand off
  • what to bet on
  • agent runtimes and harnesses
  • tool use
  • memory
  • evaluation
  • emerging class of agent concerns
  • agent identity and permissioning
  • guardrails and safety
  • cost governance for LLM and agent workloads

Other signals

  • agent platform
  • runtime and execution environment
  • memory
  • identity and permissioning
  • discovery and registry
  • evaluation
  • tool/MCP interfaces
  • paved path
  • member scale
  • capture reasoning and tool-call trajectory
  • structured, replayable data
  • real-world outcomes into evaluation signal
  • memory that compounds across runs
  • feedback loop
  • improved behavior
  • eval gates
  • safe rollout
  • first-class platform capabilities
  • measurably improve over time