Hands-on playbook for delivering software at the speed of AI

AI accelerated code generation but exposed every weakness in the delivery pipeline. CI times doubled as test volume exploded. Semantic merge conflicts break main daily. Code review cannot keep pace with agent-generated PRs.

In this 90-minute hands-on workshop, Matt Matheson, CTO and co-founder of Trunk, and Patrice Boisclair-Laberge, staff platform lead, walk through a three-part playbook using a sample GitHub repo attendees can fork and follow along with live:

• Make CI fast: Implement test selection, parallelization, and flake quarantine. See the flake math in action: 10 flaky tests at 5% failure rate cost 8 engineering hours per day.
• Keep main deployable: Configure a merge queue to prevent semantic merge conflicts that individually passing PRs create together.
• Automate code review: Layer in linters and AI review for conventions, freeing humans for intent and architecture.