| |
|
| |
| DETAILS |
|
(You must register through AWS platform to be able to attend the event - click here https://events.builder.aws.com/event/1f856457-4c93-42ca-ad7e-de4784de0fcc/register?Refid=sf-loft-cal-public)
From hallucination evaluation to enforceable decision control : for teams running agents beyond demos.
Join us at the AWS Builder Loft in San Francisco for a technical session designed for teams already testing or running Amazon Bedrock Agents in real workflows.
Most agent demos work.
Production is where things break, not because models are weak, but because agent decisions are implicit, non-deterministic, & hard to audit.
Rippletide ships a hallucination evaluation module today, used by teams to detect uncertainty & risky outputs in Bedrock agent workflows. But evaluation alone doesn't make agents deployable.
In this session, we'll show how teams use Rippletide to turn those signals into production-grade control:
block or rewrite unsafe actions before execution
apply deterministic policies across agents & models
generate a full audit trail of every decision & outcome
You'll see concrete patterns, a live walkthrough, & how Rippletide extends Bedrock without changing your existing agent setup.
This session is not an introduction to agents or LLMs. It's for teams moving from prototype to production, who need governance, traceability, & confidence to scale.
If you're running Bedrock Agents & don't fully trust their decisions yet, this session is for you.
Limited seats. Priority given to teams actively deploying Bedrock Agents.
|
|
|
|
|
|
|
|
