AI/ML + open source observability

on YouTube ↗23m

" summary

Grafana adds generative AI features into open source observability: an LM plugin lets users bring an OpenAI key or self‑host models and vector stores, powering incident auto-summary (ChatGPT generates root cause, mitigations and impact), PromQL query advisor (natural‑language → PromQL) and a sift log explainer. Features run in Grafana Cloud and OSS, ship benchmarks and a public evaluation framework, and the live demo notes a roughly 10% nondeterministic failure rate for API calls.

AI-written summary · about

" chapters (9)

1. 0:00Introduction
2. 1:04AI/ML in observability
3. 3:15Reducing toil with AI
4. 4:19Large language models
5. 6:49Generative AI at Grafana
6. 10:23Responsible AI frameworks
7. 13:59Live demo: Grafana AI
8. 19:39Grafana Cloud AI features
9. 20:42Future of Grafana AI

" related talks

• 23mAI/ML + Grafana: How to Create Intelligent Grafana Apps Leveraging LLMsNathan Marrs · GrafanaCON 2024
• 26mAI-Driven Cloud Observability: Reduce toil and amplify time to resolutionKyle Shelton · Commit Your Code 2024
• 28mWhere are we going with observability?Adrian Cole · DataEngBytes Sydney 2024
• 21mAgent Observability Beyond VibesIvana Huckova, Sven Großmann · Agent Conf 2026
• 38mEnhanced Log Analytics Using Generative AIGyula Halmos · Compass AI Session 2024
• 25mAI-Powered O11y in Microservices ArchitecturesSrinivas Sriram Mantrala · Conf42 Observability 2025
• 30mAI Agents for Incident ManagementBirol Yildiz · SREday Cologne 2025 Q2
• 33mWhere Are We Going With Observability?Adrian Cole · KubeDay Australia 2024