Talks and executive sessions on agentic AI SDLC

Code is no longer written. It is orchestrated.

A session for your engineering teams: a keynote on how the work of a developer is changing, a live demo of the practices that work today, and an open Q&A. Before the session I meet with your leaders to shape it around what matters most to the company.

The state of the art in AI engineering

How the software lifecycle has changed, step by step. For each step, what the leading companies do, where they agree, where they disagree, and where the industry is going.

• Why review becomes the new queue: on teams that use AI the most, developers merge 98% more pull requests and review time grows 91% (Faros AI, 2025)
• The harness around each project
• The corporate brain that gives agents the knowledge of the company
• The infrastructure agents run on: cloud devboxes and ephemeral test environments for every PR
• How to split inference between frontier models and self-hosted models
• The software factory that runs the work

AI engineering: governance and scale

What a company needs in place before agents run at scale, the controls that leading companies use today, and how to roll them out without slowing the teams down.

• Sandboxes that limit what an agent can touch
• Credentials that agents never see in clear text
• An audit trail of every action an agent takes
• Cost control, so the bill grows with the value and not faster