AWS AI League: Cisco Upskills 1,200 Engineers in Agentic AI Through Competition
AWS AI League, used at Cisco's Transformation Fest, provides engineers with hands-on AI development experience. Through gamified challenges involving model customization and agent building with Amazon Bedrock AgentCore, Cisco enabled large-scale practical upskilling in generative AI.
What changed?
Amazon Web Services (AWS) introduced AWS AI League, a gamified, hands-on AI learning platform, to Cisco at scale. In 2026, 1,200 Cisco engineers took part in an AWS AI League competition as part of Cisco Transformation Fest. Participants competed in two tracks: building agentic AI systems using Amazon Bedrock AgentCore and model customization using Amazon SageMaker Studio. Activities included real-world tasks such as content handling and data extraction. Engineers received automated feedback via a live leaderboard and finalists competed in a game-show-style event.

Why does it matter to an everyday developer?
Traditional AI upskilling programs often involve passive content. AWS AI League emphasizes active learning by engaging engineers in practical, competitive challenges mapped to real business problems. Developers gain firsthand skills in building, fine-tuning, and evaluating AI models and agent systems. The format encourages experimentation, iteration, and peer comparison, helping engineers rapidly develop competence in current AWS AI tools. This approach can foster greater confidence and productivity in deploying generative and agentic AI solutions.
What can the developer do now?
Developers interested in advancing their AI skills on AWS can participate in upcoming AWS AI League events, either internally in their organizations or at public AWS-hosted competitions. They can leverage resources like Amazon Bedrock AgentCore for agentic AI development and Amazon SageMaker Studio for model customization. Organizations can also host similar competitions to accelerate practical AI adoption. Practical steps include attending workshops, experimenting with challenge platforms, and reviewing past challenges for learning pathways.
