AWS commits USD $1 billion to embedded AI engineers

Amazon Web Services has created a Forward Deployed Engineering organisation and committed USD $1 billion to it. The unit is intended to place AI engineers inside customer teams.

The new group is designed to help customers build and deploy agentic AI systems directly within their operations, rather than through a conventional consulting model. Its engineers will work with the customer's business, engineering, and security teams and aim to move projects from development to production in days rather than months.

The move reflects a shift in demand from companies that have already tested AI tools and now want to embed them more deeply into business processes. These customers are looking for help redesigning workflows around AI systems and developing internal teams that can continue operating those systems after an engagement ends.

Under the model, AWS plans to embed what it calls frontier teams within customer organisations. These engineers will use AI agents alongside human oversight to build systems that leverage customer data, governance frameworks, and internal processes.

Delivery model

AWS positioned the organisation as distinct from project-based advisory work by tying engagements to shared business goals rather than to billable hours. Customers are meant to leave with both deployed software and the technical skills, workflows and documentation needed to run and extend those systems themselves.

That includes knowledge graphs, runbooks, architectural records and internal staff trained to operate independently. AWS said the semantic layer deployed during these engagements sits inside the customer's own AWS account, connects to enterprise data sources and produces a governed, versioned knowledge graph for AI agents to use.

The company highlighted security features including hardware-based isolation, end-to-end encryption and retention of customer data within the customer's own governance framework. AWS said these controls are built into deployments from the start, an issue likely to be closely watched in regulated sectors, including financial services and government.

Early customers

Engineers from the new organisation are already working with customers including the Allen Institute, Cox Automotive, the NBA, Ricoh, Southwest Airlines and the NFL. AWS used the NFL as its main example of how the model has been applied in practice.

Gary Brantley, Chief Information Officer at the National Football League, described the work in a statement from AWS. "The NFL has millions of fans who want to consume football content throughout the year, including the offseason. We innovate at the pace and scale needed to meet the high expectations of our fans," Brantley said.

"To create new digital experiences for our fans, the NFL partnered with AWS FDE and got engineers building alongside our team to launch into production in just weeks. 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," he said.

Broader push

The investment builds on AWS's broader effort to turn interest in generative AI into production workloads running on its cloud platform. The company said it has been developing AI solutions for customers since 2017 and that, over the past three years, engineers in its Generative AI Innovation Centre have worked on thousands of customer projects.

AWS cited earlier work with BMW, Jabil and Lyft as examples of that experience. According to the company, those engagements ranged from reducing service disruptions across connected vehicles to developing a manufacturing assistant and improving driver support response times.

The new organisation also has implications for AWS's partner ecosystem. Partners will contribute model expertise, industry knowledge and other skills, and AWS said it is investing in partner training, tools and resources tied to the new engineering model.

Competitive context

The announcement comes as major cloud providers compete to demonstrate not only that they can supply AI infrastructure and models, but also that they can help customers deploy systems quickly and safely. Large organisations have often struggled to move from pilots to operational deployments due to integration work, governance concerns, and shortages of in-house engineering talent.

By offering embedded engineering teams, AWS is seeking to address those barriers while deepening customer reliance on its platform. At the same time, it is trying to reassure clients that the end result will not be long-term dependence on external specialists, but internal teams that can maintain and develop AI systems on their own.

The offering is aimed at organisations that have moved beyond experimentation and need production AI systems for real business processes, particularly in sectors where security, governance and speed to production are tightly scrutinised.