Beyond Code Generation: Governing AI Agents Across the DevSecOps Lifecycle
summary
Argues that governing AI agents across planning, coding, security and CI/CD avoids shifting the bottleneck downstream and can turn weeks of triage into minutes: a repo with 51 high vulnerabilities goes from weeks of manual prioritization to automated prioritization and issue creation in minutes. Describes building an orchestration layer using GitLab Duo, planner, security and data analyst agents, a shared graph context and agent-to-agent protocol. Demonstrates an auto‑generated tic‑tac‑toe app and automated MR pipeline that largely works but needs guardrails and human review.
AI-written summary · about
chapters (7)
- 0:00AI agents in DevSecOps
- 7:47DevSecOps platform journey
- 14:34AI productivity paradox
- 20:29Intelligent AI orchestration
- 27:18Enterprise AI collaboration
- 29:51Live demo: AI agents in GitLab
- 49:59Q&A
related talks
35mManaging the AI Surge: Governing & Scaling in the AI Era
30mBuilding Tools with Tools: Modern AI-Driven Development
31mAgents at Work: How GitLab Is Redefining AI Adoption for Engineering Teams
19mIf AI Writes the Code, Who Runs Production?
33mThe Autonomous Enterprise
18mUsing AI Across the Dev Lifecycle
34mBeyond CodeGen: How AI & Machine Learning Are Transforming DevOps
58mSecuring Code in the Age of AI