Google SRE Prodcast The One With Denia Del Cid and AI
Jan 21, 2026
Denia Del Cid, an SRE at Google who applies AI to reliability problems. She talks about early outage detection, incident similarity analysis, and cutting repetitive toil with LLM-powered tools. Denia stresses validating models with golden datasets and keeping humans in the loop. She also explores agent-driven on-demand analysis and practical steps for building trust in AI systems.
AI Snips
Chapters
Transcript
Episode notes
AI For SRE Is Rapidly Maturing
- Denia Del Cid sees AI for SRE as a rapidly evolving, practical field after a year and a half of focused work.
- Model improvements (Gemini versions) changed what tasks AI could reliably help with in production workflows.
Early Outage Detection Proved The Concept
- Denia described a proof-of-concept that analyzed customer support escalations and detected outages earlier than traditional automation.
- That early detection convinced teams the approach could surface meaningful production signals.
From Weeks Of Analysis To Minutes
- Denia's tooling reduced weeks of manual ticket analysis to minutes and produced dynamic dashboards for planning.
- That shift freed SREs to focus on higher-value design and coding work instead of manual tagging.