The Agent That Turned On-Call Chaos into a Single Slack Thread
summary
Describes wiring Datadog, PagerDuty, CloudWatch, Slack, and Linear into Retool Agents and custom Python tools to turn on‑call alert noise into a single Slack thread. The system correlates alerts, metrics, and AWS events into one incident timeline, runs lightweight Python analysis to suggest probable causes, and posts actionable Slack buttons (one‑click rollbacks, workflows) that trigger real rollbacks and create synced Linear followups.
AI-written summary · about
chapters (12)
- 0:00On-call chaos in Slack
- 2:34Automate incident response
- 4:23Agent vs scripts
- 6:06Live demo: incident agent
- 9:00Building agents in Retour
- 10:05Agents with natural language
- 12:55Agent tools and resources
- 15:45Root cause analyzer
- 17:30Slack and Linear integration
- 18:34Triggering agents
- 20:41Reducing on-call burden
- 21:44Q&A
related talks
24mOn-Call the Easy Way With Agents
34mIncidents Happen
16mHuman Centered On Call Practices
31mOn the Importance of Visibility
24mWait for Us! Evolving On-Call as Your Company Grows
43mSupercharge your incident response with Trek10 and Datadog
27mBuilding a Scalable Monitoring System
23mFirefighting: The Engineer's Guide to On-Call