Principal Forward Deployed Engineer

Unity Unity · Enterprise · San Francisco, CA · Engineering

Principal Forward Deployed Engineer to embed with departments (Finance, Legal, Marketing, etc.) at Unity to transform workflows using AI, agents, and automated workflows. This role involves writing production code, designing and deploying agents, building integrations, and partnering with non-technical stakeholders to solve real problems and improve efficiency.

What you'd actually do

  1. Embed with a partner department and learn their workflows, decision points, systems, and unwritten rules well enough to know where AI will earn its keep and where it won't.
  2. Design, build, and ship agents and automated workflows that solve real problems in that department — replacing manual work, unblocking bottlenecks, and surfacing information that is currently buried.
  3. Write production-grade code across the stack: integrations, services, data pipelines, agent logic, and the occasional internal UI.
  4. Connect the systems the department lives in — their SaaS tools, internal services, data warehouse, and communication channels — and build the integrations that don't exist yet.
  5. Run honest evaluations before anything touches a real workflow, and instrument what you ship so it stays trustworthy after launch.

Skills

Required

  • Production software engineering experience (services, integrations, data pipelines, APIs)
  • Hands-on experience building with LLMs and agentic systems (prompt and context engineering, tool use, retrieval, evals, agent loops, failure modes)
  • Experience shipping agentic systems used by real people
  • Strong instincts for working with non-engineering partners
  • Comfort with ambiguity
  • Bias toward building lasting things
  • Effective written and verbal communication

Nice to have

  • Prior Forward Deployed Engineer, internal tooling, or solutions engineering experience at a company doing serious AI deployment
  • Experience designing evals and observability for agentic systems in production
  • Familiarity with MCP, tool-calling frameworks, or agent orchestration platforms

What the JD emphasized

  • production code
  • ship real software
  • significant production software engineering experience
  • You ship code; you don't just design it.
  • Hands-on experience building with LLMs and agentic systems
  • You have shipped something agentic that real people used.

Other signals

  • building agents
  • production code
  • integrations
  • non-technical partners