Amazon Web Services is making a $1 billion bet that AI adoption will move faster if its own engineers move inside customers’ offices, escalating a competition over how hands-on tech giants should be in deploying artificial intelligence.
Early roots: Palantir’s model goes mainstream
The “forward-deployed engineer” (FDE) model was pioneered more than a decade ago by Palantir, sending specialists to work from within client businesses rather than from the vendor’s own offices.1 The approach promised faster, more tailored deployments of complex software systems, and has since spread across enterprise tech as AI tools became harder for companies to integrate on their own.1
OpenAI and Anthropic move first
In May 2026, AWS partners Anthropic and OpenAI each launched their own FDE-style joint ventures with major private equity backers to help enterprises roll out their AI models.2 Anthropic teamed up with firms including Blackstone, Hellman & Friedman, and Goldman Sachs to target mid-sized companies, while OpenAI set up a similar deployment company with investors such as TPG, Advent International, Bain Capital, and Brookfield.2 These ventures framed embedded AI services as central to the race to sell enterprise AI.
AWS enters with its own $1bn FDE unit
On June 30, 2026, AWS announced a new internal organization for AI-focused forward-deployed engineers, committing $1 billion in internal resources rather than outside capital.1 The unit will start with “thousands” of engineers organized into pods of five or six, each embedded with a single customer at a time to help build and run AI systems and agents.2
Francessca Vasquez, AWS vice-president of frontier AI engineering and services, said the pitch “came down to one word: speed,” arguing that customers want quick returns and that FDE pods will hand back “a self-sufficient team within weeks.”2 AWS also emphasizes that customers will leave deployments not just with “agentic systems” but with “lasting AI skills, workflows, and patterns they can use to innovate independently.”1
Competing visions, shared goal
Unlike OpenAI and Anthropic’s investor-backed joint ventures, AWS is folding the effort into a single business unit, which Vasquez described as “the first time we’re doing it in that way” despite existing in-house expertise.2 All three players, however, are converging on the same idea: companies need embedded experts, not just cloud credits and APIs, to turn AI hype into working systems at scale.12