The initiative will rely on so-called forward-deployed engineers—specialist technical staff who work closely with client teams on-site or virtually, helping integrate and operationalize AI systems into real business workflows. The company is committing an initial $1 billion to the programme.
According to Francessca Vasquez, AWS vice president of frontier AI engineering and services, the plan involves deploying small groups of engineers in structured “pods” for short, intensive engagements lasting about 45 days at a time.
“We have a ton of demand for customers who are asking for our help to really drive agentic AI patterns in their workflows,” Vasquez said in an interview prior to the announcement.
She explained that the goal is to move beyond traditional consulting models by embedding engineers directly with customers, where they can navigate internal processes, build production-ready systems, and help translate AI capabilities into working applications.
Forward-deployed engineers are not a new concept in the broader tech industry. Companies like Palantir Technologies have used similar embedded engineering models for more than a decade. Other firms, including Salesforce, Anthropic, and Google Cloud, have also expanded customer-facing engineering support structures in recent years as demand for enterprise AI integration grows.
Industry interest in this type of role has surged alongside the AI boom. Forward-deployed engineering is increasingly viewed as a hybrid between software engineering and consulting, with professionals expected to not only build systems but also understand client operations deeply enough to adapt solutions in real time.
AWS says it intends to scale the new unit to “thousands” of employees over time, combining internal transfers with external hiring. The company is also positioning the initiative as a rare area of expansion amid broader workforce reductions, noting that Amazon has cut more than 30,000 corporate roles since October.
The new division is being launched during a major AWS customer event in Washington, where additional announcements are expected, particularly around government-focused cloud services.
Early clients already include organizations such as the National Basketball Association and electronics firm Ricoh.
Success for the programme will be measured by how quickly customers can develop new products or upskill their teams using AWS support, according to Vasquez. “We want to make sure that these customers get value in faster durations than what they’ve traditionally seen in project-based activity,” she said.
The move underscores Amazon’s effort to compete more aggressively in the fast-evolving enterprise AI market, where speed of implementation and hands-on support are becoming as important as raw computing power.