AWS announced a new $1 billion organization called Forward Deployed Engineering (FDE) at AWS Summit D.C. on June 30. The unit will embed thousands of AWS AI engineers directly inside customer organizations to co-develop and deploy agentic AI solutions, according to Francessca Vasquez, VP of Frontier AI Engineering and Services at AWS.
The program targets companies that have moved past AI experimentation and need production systems running real business processes. The NFL, Southwest Airlines, Cox Automotive, the Allen Institute, Ricoh, and the NBA are already working with FDE teams.
How FDE Deployments Work
FDE engineers embed directly alongside a customer’s business, engineering, and security teams. According to AWS, the teams use what AWS calls the “AI-Driven Development Lifecycle,” a methodology that combines AI-powered execution with human oversight. Each customer deployment compounds intelligence for the next project.
The pitch is speed. AWS claims FDE compresses AI deployment timelines from months to days. That claim got a concrete example from the NFL. Gary Brantley, the NFL’s CIO, said the league partnered with FDE and “got engineers building alongside our team to launch into production in just weeks,” producing fan-facing products like NFL Fantasy AI and NFL IQ, according to AWS.
Not Consulting, Says AWS
AWS is positioning FDE as structurally different from traditional consulting. Vasquez wrote that deployments are “structured around shared goals and business results, not billable hours.” Each engagement is designed to leave the customer self-sufficient: customers get deployed agentic systems, knowledge graphs, runbooks, architectural documentation, and trained internal teams.
At the center of each deployment is a semantic layer that connects to enterprise data sources and publishes a governed, versioned knowledge graph. Agents reason over that knowledge graph, so institutional expertise persists in the customer’s code rather than in individual engineers who might rotate off the project.
The organization builds on three years of work by the AWS Generative AI Innovation Center, which has delivered thousands of customer AI solutions. Prior engagements include BMW (reducing service disruptions across 23 million connected vehicles), Jabil (manufacturing floor assistant), and Lyft (resolving driver support issues 87% faster).
The Consulting Industry Question
The $1 billion figure and the “not billable hours” framing position FDE as a direct challenge to firms like Accenture, Deloitte, and McKinsey, all of which have built large AI consulting practices. If AWS can deliver production agentic systems in weeks at the scale Vasquez describes, the value proposition for paying consulting rates on multi-month AI engagements gets harder to justify.
The counterargument is obvious: AWS has a massive infrastructure business to feed. Every FDE engagement produces systems running in the customer’s AWS environment. The engineers leave, the compute bills stay.
For organizations building autonomous agent systems, the announced customer list signals where FDE sees immediate demand: sports media (NFL, NBA), automotive (Cox Automotive), airlines (Southwest), manufacturing (Ricoh), and scientific research (Allen Institute). Regulated industries, financial services, and government are explicitly called out as target verticals.