Tech Week - Agentic AI in production: retrieval, drift, and what actually works

We’re thrilled to unite the Elastic, and Chainguard communities for an evening of learning and connection.

Join us to learn something new and meet others who are passionate about search, observability, and security.

Thursday, May 28th

4 Longfellow Pl. Boston, MA 02114

Agenda:

5:30 pm: Doors open; say hi and eat some food6:00 pm: DriftlessAF: An open source agentic reconciliation framework proven at scale, by Maxime Gréau, Principal Software Engineer at Chainguard6:30 pm: Q&A with Maxime6:40 pm: The Agent That Retrieves Best, Wins, by AJ Meyers, Principal Solutions Architect at Elastic7:10 pm: Q&A7:20 - 7:30 pm: Networking and event wrap-up

Abstracts:

DriftlessAF: An open source agentic reconciliation framework proven at scale

At Chainguard, we build and maintain over 2,000 unique containers, hundreds of thousands of package versions, and hundreds of CVE patch backports. Our old event-driven architecture couldn't keep up: brittle queues, duplicate failures, lost work items. We had to start over.

The answer was DriftlessAF, our open source agentic reconciliation framework. A fleet of AI agents continuously watches for drift between the real world and the desired state, decides what to do about it, and acts.

In this talk I'll share what it actually takes to run agentic workflows reliably in production: how we design agents that evaluate their own work, and what we got wrong before we got it right. I'll demo DriftlessAF live, with Elastic powering the observability layer so you can see exactly what our agents are doing under the hood.

Bio

Maxime Gréau is a Principal Software Engineer at Chainguard, where he's part of the Platform team building DriftlessAF, an open source agentic reconciliation framework that powers Chainguard Factory 2.0. Previously at Elastic for over seven years, he led cross-engineering supply chain security efforts that reduced CVEs across Elastic's container images, pioneered Sigstore container signing, and contributed to the design of GitOps-based serverless CD workflows. He's a published author ("Apache Maven") and a conference speaker.

The Agent That Retrieves Best, Wins

Enterprise AI agents are only as intelligent as what they retrieve — yet most organizations are investing heavily in models while ignoring the layer that determines whether those models reason accurately, act on current knowledge, and deliver distinctive results. This session reveals why retrieval is the decisive competitive variable in agentic AI, exposes the organizational and architectural gaps most teams don't know they have, and gives every attendee a concrete framework to assess where they stand — and what to do Monday morning.

Bio

AJ Meyers has been a Principal Solutions Architect at Elastic for the past seven years. He helps customers adopt Elasticsearch into their operations, security, and application development workflows. Prior to Elastic, AJ brings over three decades of extensive experience delivering data-forward strategies, infrastructures, and solutions holding senior architectural and leadership roles at major technology and consulting firms including Teradata, FICO, Deloitte Digital, and E&Y

See you there!