AWS bets $1 billion on engineers who build AI systems inside client teams
Amazon Web Services is investing $1 billion in a new unit that will place thousands of AI engineers right inside enterprise teams.
The goal is to shrink agentic AI rollouts from months down to days.
What Forward Deployed Engineering actually builds
AWS Forward Deployed Engineering, or FDE, is the unit’s name. It matches the business, engineering, and security personnel of a customer with AWS engineers. Together, they use the customer’s data and governance guidelines to create production AI systems.
The program was announced by Francessca Vasquez, vice president of frontier AI engineering and services at AWS.
FDE skips the usual consulting playbook, with no assessments or slide decks of recommendations. Engineers build production systems alongside the customer’s own people, and the work centers on shared business outcomes rather than billable hours.
According to AWS, it wants clients to be independent after the engagement is over. Customer engineers begin by observing, progress to co-building, and ultimately manage systems independently. Documentation, runbooks, knowledge graphs, and personnel trained to function without AWS in the room are all left behind after every deployment.
At the technical core, a semantic layer is installed within the customer’s AWS account. It publishes a governed knowledge graph that AI agents can reason over, connects to enterprise data sources, and enhances metadata. Instead of locking in individuals who might leave, AWS wants domain expertise integrated into code and systems.
FDE teams already count the NFL, NBA, Allen Institute, Cox Automotive, Ricoh and Southwest Airlines among partners. NFL Fantasy AI, NFL IQ – those are fan-facing products that have come out of the partnership, according to NFL Chief Information Officer Gary Brantley.
Brantley stated, “Together, we created new fan-facing products like NFL Fantasy AI and NFL IQ that allow fans to interact with NFL data like never before. The engagement from fans and broadcasters was measurable from day one and was made possible by AWS’s delivery model.”
AWS says its engineers were able to get the NFL’s production systems back online within a few weeks.
AWS’s FDE builds on three years of AI deployment work
FDE did not appear out of thin air. According to Vasquez, AWS’s Generative AI Innovation Center has worked on thousands of customer AI projects over the past three years. The group assisted BMW in minimizing service interruptions for 23 million connected cars. It built an assistant for manufacturer Jabil on the factory floor. With Lyft, it reduced driver support resolution times by 87%.
For clients who desire more extensive AI capabilities across operations rather than just one-time use cases, FDE extends that model. AWS says it is investing in partner training and tools to support the program, and its partners will also contribute model expertise and industry knowledge.
The $1 billion pledge coincides with additional AWS infrastructure initiatives. According to Data Center Knowledge, the company launched its Graviton5-powered EC2 C9g and C9gd instances generally with a focus on distributed analytics workloads, high-performance computing, and AI orchestration. The compute layer that powers AI agents is managed by Graviton5.
AWS is deploying FDE to companies that have gone through the testing phase. The target list leans towards government, financial services and regulated industries, where security and governance regulations are impeding the uptake of AI.
It directed businesses interested in FDE to reach out to their AWS account teams.
Don’t just read crypto news. Understand it. Subscribe to our newsletter. It's free.
Recommended Articles
