Show HN: AIOps Cockpit – An open-source AI SRE with human-in-the-loop safety

Category: observability

Tags: aiops, sre, llm-agent, human-in-the-loop, incident-response, 3d-visualization, self-hosted

Score: 5.8/10 (Innovation: 6, Technical: 7, Documentation: 5, Utility: 5)

AIOps Cockpit is an open-source AI-powered SRE control room that uses an LLM agent (Claude 3.7) to diagnose incidents by reading logs and metrics, then proposes fixes as unified diffs requiring human approval with a two-person rule. It features a 3D spatial topology map, real-time telemetry, and defense-in-depth safety mechanisms, aiming to combine AI assistance with strict operational safety. The project is notable for its human-in-the-loop design and novel 3D visualization, though it remains early-stage with limited community traction.

Target audience: devops, site reliability engineers, backend developers, infrastructure engineers

Repository: https://github.com/matricphase-dot/aiops · TypeScript · 2 stars

View on Hacker News