The FDE Playbook for AI Startups with Bob McGrew

Bob McGrew, former OpenAI CRO and Palantir executive, explains why the Forward Deployed Engineer model has become the dominant go-to-market strategy for AI agent startups.

• Over 100 YC startups are now hiring Forward Deployed Engineers, up from essentially zero three years ago.
• Shyam Sankar, now Palantir president and CTO, invented the FDE model after realizing product discovery had to happen from inside the customer, not outside via sales.
• AI agents have no incumbent product, making heterogeneous product discovery unavoidable — exactly the condition that forced Palantir to invent FDEs.
• The FDE success metric is rising contract value per customer, not falling cost per customer — the inverse of standard SaaS PMF logic.
• Echo team (domain-expert ‘heretics’) and Delta team (fast-prototype engineers) are the two non-negotiable structural roles; hiring traditional enterprise salespeople almost never worked at Palantir.
• Palantir’s famous ontology — generic objects, properties, and links rather than customer-specific schemas — emerged directly from FDE field feedback needing cross-customer generalization.
• McGrew argues AI capabilities are advancing faster than adoption, making the real opportunity filling the gap between what models can do and what enterprises can absorb — startups as FDEs to OpenAI’s ‘home product team.’

2025-09-08 · Watch on YouTube