AWS commits $1 billion to embed AI engineers inside customer teams

Amazon Web Services committed $1 billion on June 30, 2026 to a new Forward Deployed Engineering organization that embeds thousands of AI engineers directly inside customer teams to build and deploy agentic AI systems. The AWS Forward Deployed Engineering unit is led by Francessca Vasquez, vice president of frontier AI engineering and services, the company said.

The move is AWS copying a consulting-flavored playbook rather than inventing one. OpenAI and Anthropic already run forward-deployed-engineer teams, and demand for the role grew 42-fold from 2023 to 2025, according to a LinkedIn report cited by AWS. The bet is that selling cloud credits is no longer enough; customers need engineers in the room to turn AI pilots into running systems.

The model is structured for short engagements. AWS plans to send engineer pods to customers for roughly 45-day periods, and customers come away with deployed systems, knowledge graphs, runbooks and trained internal staff to keep them running, AWS said. The company is aiming the program at regulated industries, financial services and government.

Early customers include the Allen Institute, Cox Automotive, the NBA, the NFL, Ricoh and Southwest Airlines. AWS said the NFL used the program to launch its NFL Fantasy AI and NFL IQ products in just weeks, a timeline the company has not substantiated with detail.

The results so far are AWS’s own account, and a $1 billion headline spread across thousands of engineers and a global customer base is a commitment, not a measured outcome. Whether short embedded engagements produce durable systems or handoffs that stall is the open question.