AI-augmented code modernization: from framework migrations to autonomous agents (Oct. 7th)
Wednesday, October 7th, 2026: 3:00 PM to 4:00 PM
What it takes to modernize a large repository portfolio with AI agents, and what that actually costs to run.
Virtual
The Challenge
Framework migrations stall, CVE backlogs compound, and a fix that works on one repo rarely scales to hundreds. Point a general-purpose AI agent at the problem and it re-reads the codebase on every run, spends most of its tokens rediscovering what it already saw, and still returns changes someone has to review by hand. Making that repeatable across a large portfolio, without the token bill and review load climbing with every repo, is what separates a proof of concept from a rollout that actually finishes.
Inside the session:
What "done" actually means across a full portfolio: every service on the target version, changes verified, no regressions
The economics of running agents at scale: what actually drives the token bill, and the levers that bring it down as you add repositories instead of letting it balloon
Where autonomous code engineering is headed: continuous, goal-directed remediation across the codebase with engineers in the loop
A live demo: kicking off a change and watching it roll out across an enterprise codebase, with the token cost visible as it runs
Hosted by DevOps.com