OpenAI describes its forward deployed engineering team
Original title: What FDE means at OpenAI
The team, founded in early 2025, embeds with customers and later rebuilt semiconductor workflows on Codex.
OpenAI says its forward deployed engineers close the gap between model benchmarks and reliable customer use, founded as a team in early 2025. Engagements with John Deere and an unnamed semiconductor company are given as examples, including personalized support around See & Spray and AI-native chip verification. OpenAI reports production use within 3–6 months and later moved custom apps onto shared Codex primitives, retiring separate UIs. It says success means customers no longer need FDEs for that class of problem.
Unconfirmed: Time-saved and performance figures are OpenAI’s account and are not independently verified here.
- See & Spray is described as cutting chemical use by up to 70 percent when farmers configure it well.
- About 15 FDEs worked more than 10 semiconductor verification use cases in mid-2025.
- OpenAI reports about 48 percent time saved on directed test generation and 22 percent median kernel-optimization gain.
- Verification can account for about 70 percent of project effort in that industry, OpenAI says.
- Rebuilding on Codex removed roughly 90 percent of custom code from each use case.
- Products
- CodexSee & Spray
- Organizations
- OpenAIJohn Deere
From the original article
8 min read at openai.com
When Codex came out, someone on the team asked me, “What do we do when Codex can just one-shot the use cases we spent the last few months building?” That question gets to the heart of what FDE means at OpenAI. We work alongside customers to understand their hardest problems and build solutions they can depend on.
openai.com
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.