Amazon Assembles New AI Team to Accelerate Agentic AI Development
The technology giant is strengthening its artificial intelligence capabilities through a specialised research group focused on developing new AI tools and experimenting with autonomous agents.
Amazon has assembled a specialised artificial intelligence team within its cloud computing division, Amazon Web Services (AWS), to accelerate innovation in agentic AI. The group, known as Velocity Labs, was established in April 2026 and consists of approximately 35 scientists and engineers, most of whom are based in Seattle. Led by AWS Vice President Marc Brooker, the team is designed to experiment with emerging AI technologies, develop prototypes quickly and explore opportunities that traditional product-focused teams may overlook.
Unlike conventional development teams, Velocity Labs encourages engineers to participate in research while scientists contribute directly to engineering projects. The approach is intended to reduce development time and help Amazon respond more quickly to technological breakthroughs. The team demonstrated this strategy in October when it developed and publicly released an open-source AI model called Strands Decider 2B in less than a week. The model is designed to help AI agents make quick decisions, including selecting appropriate tools and directing requests to the right systems.
The development also reflects a major transformation in Amazon’s software engineering operations. Brooker revealed that AI agents now generate almost all implementation code for the vast majority of the company’s developers, compared with just two years ago when most programmers still wrote their code manually. Amazon is consequently updating its engineering guidelines to encourage employees to adopt AI tools while prioritising actual business results rather than measuring productivity through lines of code or the number of AI interactions.
The expansion comes as Amazon increases its investment in artificial intelligence amid intensifying competition with Microsoft, Google, OpenAI and Anthropic. However, the company’s rapid AI transition is taking place alongside significant workforce restructuring, with approximately 30,000 corporate jobs cut since late 2025. Amazon says it wants engineers who can adapt to new technologies, take responsibility for their work and solve complex problems. Velocity Labs could serve as a model for additional specialised teams as the company seeks to strengthen its position in the rapidly evolving AI industry.
