← Forward Deployed Engineer
Forward Deployed Engineer

Research Spotlight: What the Postings Actually Say, Compared

16 Nov 2026·5 min read

Secondhand summaries of "what companies want" drift from what a posting actually says, one retelling at a time. This series has tried to model a different habit throughout: read the primary document yourself. Here is what that habit turns up when applied to this specific role, across all three companies that currently use the title on their own career pages.

What is common across all three

  • Strong fundamentals over credentials — none of the three postings gate the role on a specific degree or a strict years-of-experience number as the headline requirement.
  • Customer-facing communication is listed as a stated requirement in every posting, not folded in as a vague "soft skill" afterthought.
  • Comfort owning ambiguity — language like "minimal supervision" (Palantir) and "own discovery, scoping, design, build and production rollout" (OpenAI) appears across the postings in different words but the same shape.

What differs

OpenAI's postings are explicit that success is measured by production adoption and eval-driven feedback — a stated metric, not just a description of the day-to-day work. Palantir's framing leans harder on autonomy and the "startup CTO" comparison. Anthropic's own Greenhouse board lists both individual-contributor and leadership tracks ("Head of Forward Deployed Engineering"), which is itself evidence this has become a discipline with a management path, not just a single job title.

Do this in 60 secondsFind a job posting for any role you are interested in, at any company. Read only the posting itself end to end — not a summary of it anywhere else, not someone's LinkedIn post about it. Write down the one requirement in it you are currently weakest on, and one concrete thing you could do this month to close that gap.
Today’s AI angleAll three companies in this comparison are AI companies (or, for Palantir, deeply AI-invested), which is itself telling — the forward-deployed model keeps reappearing specifically where the gap between a general-purpose AI system and one customer's messy reality needs a human to close it.
Interview prepWhy do you think multiple top AI companies independently converged on the same unusual job title? Because the underlying problem is the same across all of them: a powerful, general system still needs someone to adapt it to one customer's specific, messy environment, and that work does not fit neatly into either a sales role or a pure engineering role — so a distinct title emerged to describe it.
Research Spotlight: What the Postings Actually Say, Compared · Ali Huzaifa