From Static Rules To Reasoning Platforms: Scaling Intelligent Canary Delivery in 2026
summary
Demonstrates an AI-driven canary pipeline that replaces static rules with a reasoning layer, wiring OpenAI/GPT-style LMs into Argo Rollouts and a Java Quarkus agent to analyze telemetry (logs, traces, metrics) and autonomously block or progress rollouts. Demo shows the system creating a pull request for an NPE fix and opening an issue for a reproduced memory leak detected, preventing a bad canary from reaching production. Uses OpenTelemetry, Argo CD, vector stores for prompts, and succeeds at automated triage in the demo.
AI-written summary · about
chapters (10)
- 0:00Introduction
- 2:34Why reasoning platforms now
- 8:27Static rules to AI reasoning
- 11:37AI agents as partners
- 16:41Demo setup and architecture
- 23:13Live AI canary deployment
- 28:36AI in internal systems
- 29:38Guardrails for safe AI
- 32:29Modernize CI/CD pipelines
- 33:54Q&A
related talks
26mBeyond Argo Rollouts: Boosting Developer Experience With Intelligen...
31mAIOps: (near) Zero-Touch Production Rollout Fixes
10mLightning Talk: Open Source Canary Deployments with Application Metric Analysis
01h 04mThe Fine Art of Canary Deployments
44mAgentic Delivery, Guardrailed: AI in CI/CD & Platform Engineering (Open Source Patterns)
24mFix Production Rollouts on the Fly With Agentic AIOps
40mProgressive Delivery with Kubernetes and Argo Rollouts
24mRollout on Reception: Progressive Delivery for GenAI Apps Using Real...