What makes an FDE

The most in-demand engineer of 2026 doesn’t have the job title you’d guess.

Forward deployed engineer roles are among the fastest-growing in AI — and having watched this profile emerge across the industry and inside AWS, I think the shape of the role is widely misunderstood.

It’s not “great coder who travels.” The public AWS job descriptions say it plainly: roughly 60% building, 40% leading. The rare skill is the combination — someone who can whiteboard a business problem with a COO at 9am and have agents shipping tested code by 3pm.

Three traits I’ve seen separate great FDEs from good engineers:

1. Comfort with ambiguity. There is no spec. Writing the spec — with the customer, in their language — is the job.

2. Production instinct. Demos are easy now; agents can build a prototype in hours. Knowing what breaks at scale, under audit, inside a regulated environment — that’s the craft.

3. Product sense. Every engagement is field research. The best FDEs feed what they learn back into the platform.

My hiring advice for years has been: take full-stack engineers who are curious about ML over specialists who’ve never shipped a product. You can teach the models. You can’t teach caring about the user.

FDE is that advice, turned into a discipline.