Forward-deployed engineers: an old consulting model becomes tech's newest fad

Forward-deployed engineers (FDEs) are among corporate tech's loudest terms this year, but the practice itself is much older. "We have been doing FDE for a long time," Ryan Sheehan, a senior vice president at the $16 billion integrator SHI, told The Register.

In practice, that means sending top consultants to a client without tying them to a project, with a mission to build something valuable with new technology and the freedom to move fast.

A French restaurant, then Palantir

The term traces back to Palantir. CTO Shyam Sankar claims he coined "forward deployed engineering" in 2007, after CEO Alex Karp asked him why French restaurants are so good. Karp's answer: the wait staff is part of the kitchen staff, understanding the food and the craft rather than just carrying dishes to the table. Karp then asked Sankar to build "that, but for engineering."

On Substack, Sankar wrote that FDEs "embed alongside our customers and work to ensure our software solves their problem and not some proxy for their problem." He called them crazy enough to board a last-minute plane to Iraq and smart enough to ship quality code the same day. Investors first balked because the costs ate into margins, but Sankar argues it pays off.

AWS and Microsoft open their checkbooks

The rest of the industry is following the recipe. In June, AWS announced a dedicated FDE organization backed by $1 billion; in July, Microsoft answered with a $2.5 billion program called "Frontier Company" that it says "goes beyond" FDEs. Both pitch embedded consultants as the way to move AI experiments into production.

The vendors also admit it is not new. AWS told The Register it has offered something similar for years, starting with a 2017 "ML Solutions Lab" that loaned data scientists to customers. Taimur Rashid, a managing director at AWS, said the company now "forward-deploys" specialists who join a customer's workflows "for a defined time," with no fixed scope or hourly billing. Cisco's Carlos Pereira said his company has done this for years too, and AI makes embedded experts more valuable.

Useful, with strings attached

Analysts see real value and real risks. Gartner's March guidance warns of "a high degree of vendor dependency, security/data risk exposure, technical debt, and talent atrophy." Forrester describes a "bespoke trap" in which tailored solutions deepen vendor dependence. Omdia's Jay McBain predicts the FDE investment will "fall back into the channel" by 2031, with 82 percent of partners saying they cannot grow AI services quickly.

The open question: will the industry someday laugh at how Palantir reheated an old form of tech services under a new name, and everyone followed the recipe?