Product Manager - Agents

Applied Intuition Applied Intuition · Robotics · Sunnyvale, CA · Infrastructure, App & AI Platform (Dana)

Product Manager for an agent platform (Dana) focused on physical AI industries. The role involves owning the end-to-end product lifecycle of agents, including their creation, execution, evaluation, governance, and improvement. This includes working closely with engineers on technical tradeoffs, making build/buy/partner decisions for the agent stack, and prototyping on the platform. The role requires hands-on experience with agents, fluency in the current agent stack (tool use, sandboxing, evals, serving, context, memory), and a strong understanding of the agent landscape.

What you'd actually do

  1. Own the agent platform product end to end, from how agents are created to how they're trusted in production
  2. Work with engineers daily on what is built and how it's built, and own the product and technical tradeoffs
  3. Make build/buy/partner calls across the agent stack (harnesses, sandboxing, evals, serving) and defend them with evidence
  4. Prototype and build on the platform yourself
  5. Track the agent landscape (labs, infra startups, open source) and keep our stack choices ahead of it

Skills

Required

  • Bachelor's Degree or higher in a relevant field
  • Hands-on experience with agents: shipped an agent product, built harnesses or eval loops, or founded in the space
  • Fluency in the current agent stack: tool use and MCP, sandboxing, evals, serving, context and memory
  • Deep, current knowledge of the agent landscape and where it's going
  • Technical competence to make tradeoffs with engineers and prototype within the platform using AI tools
  • A record of shipping fast

Nice to have

  • Founder or founding-team experience, ideally in agent infrastructure
  • Public technical work such as open source projects, demos, or technical writing
  • Experience building products for enterprise or safety-critical environments

What the JD emphasized

  • Own the agent platform product end to end
  • own the product and technical tradeoffs
  • Make build/buy/partner calls
  • Prototype and build on the platform yourself
  • Hands-on experience with agents
  • shipped an agent product
  • built harnesses or eval loops
  • Fluency in the current agent stack
  • tool use
  • sandboxing
  • evals
  • serving
  • context and memory
  • Deep, current knowledge of the agent landscape
  • Technical competence to make tradeoffs with engineers
  • prototype within the platform
  • record of shipping fast

Other signals

  • agent platform
  • agent engineering
  • evals
  • model serving
  • memory