ACD305

Guarding the Generative Path: Architecting Responsible AI with Amazon Bedrock Guardrails

Room: Room 5 | Time: 11:00

Generative AI can accelerate innovation, but only when every response respects safety, privacy and trust. In this session, we step inside Amazon Bedrock Guardrails and see how purpose-built safeguards turn risky prompts, hallucinations and policy violations into confident, production-ready answers.

Let’s explore how versioned “guardrails-as-code” enables a single CloudFormation push to raise the safety bar without touching app logic, with real-time metrics and audit logs streaming straight into CloudWatch to spotlight every blocked prompt the moment it happens.

Guilherme Dalla Rosa
CTO @ MerCloud