From theory to delivery: How Atos upskilled 400 engineers in agentic AI
When Atos set out to upskill 400 engineers in agentic AI, hands-on learning was the missing ingredient. Over three days, engineers built multi-agent systems on AWS through an AI League event. This post explains why Atos chose the format, what engineers built and learned, and what other enterprises should consider.
Atos aimed to upskill 400 engineers in agentic AI, moving from theoretical knowledge to practical application. The team recognized that online courses and classroom instruction lacked the hands-on experience necessary for effective AI implementation. By partnering with AWS, Atos discovered that hands-on learning was the key to successful AI enablement.
In 2026, they organized an Agentic AI League event for the 400 engineers, offering a three-day format to build their skills. The engineers participated in a wide range of skill levels, from those with existing AWS experience to newcomers and less technical roles such as product owners and project managers. Only 5% of participants had no prior knowledge of agentic AI, while 25% had basic awareness, 50% had understanding but no hands-on experience, and 20% had practical experience with agentic AI services.
The AI League format proved advantageous over conventional training, offering immediate practical application, competitive motivation, real AWS services, and measurable outcomes. The event utilized native AWS services including Amazon Bedrock, Amazon Bedrock AgentCore, AWS Lambda, Kiro, and Amazon SageMaker. Engineers built an autonomous AI agent that navigated a dungeon maze, solving challenges, collecting coins, and reaching a treasure within a time limit and limited lives.
The scoring model rewarded successful challenge completion, coin collection, map completion, and life retention. Engineers were also rewarded for fine-tuning specialist small language models.
Written by urgent.news from AWS Machine Learning's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
