As enterprises move agents from experimentation into production, teams hit the same problems: agent sprawl with no reuse, shadow AI, no cost attribution, no unified tracing, non-standard integration, and unclear ownership after go-live. Scaling across business units also means deciding who can access enterprise data, which tools agents can use, how their behaviour is evaluated, and who is accountable when something goes wrong.
The answer is a shared agentic AI platform: an enterprise foundation for building, governing, observing and scaling agents safely, reusing enterprise controls instead of rebuilding them for every agent. This workshop shows how to stand that platform up on AWS with Amazon Bedrock AgentCore, then operate real workloads on it, with consistent platform controls and the flexibility for business units to build agents for their own needs.
What participants leave with:
Who should attend: AI engineers, platform engineers and architects, solution and cloud architects, and technical leads and engineering managers. The session is hands-on and code-focused, familiarity with Python helps.