Production software keeps breaking and it will only get worse
summary
Claims an autonomous troubleshooting product, Traversal.ai, that combines causal machine learning with LLM reasoning and agent swarms to search petabytes of telemetry and trillions of logs in parallel, surface root-cause candidates with confidence scores, and link them to specific deploys or PRs. Reported customer result: ~40% mean-time-to-resolution reduction for DigitalOcean and typical findings returned in about five minutes; integrates with Grafana/DataDog/Splunk and produces an interactive impact map for followups.
AI-written summary · about
chapters (6)
- 0:00Introduction: The Three Pillars of Software Engineering
- 2:10The Worsening Problem of Troubleshooting
- 4:15Why Current AI/ML Solutions are Failing
- 7:08Traversal.ai's Novel Approach to Autonomous Troubleshooting
- 11:35Case Study: How Traversal.ai helped Digital Ocean
- 16:03The Broader Vision for Traversal.ai
related talks
01h 00mData is out, knowing is in: Applying AI to automate cloud operations
19mFrom Alerts to Action: AI-Powered Incident Response Systems
40mWhen Incidents Fix Themselves: AI SRE in action
24mListen to Production the Way It Deserves
11mTransforming Problem-Solving with Data-Driven Insights
05mAI-Driven Incident Resolution – Hype or Reality
13mBeyond Manual Firefighting
46mLeveraging AI and Automation in DevOps