Aurora by Arvo AI is an open-source, AI-powered incident management and root cause analysis platform built for Site Reliability Engineers and DevOps teams. It uses LangGraph-orchestrated LLM agents to autonomously investigate cloud incidents across major providers like AWS, Azure, and GCP, as well as Kubernetes. The platform integrates with 22+ tools including PagerDuty, Datadog, and Slack, and is available for self-hosting or fully hosted use.
Funding
Funding not disclosed
Founders
Product
Problem
Site Reliability Engineers face significant delays in identifying root causes of cloud incidents, often spending hours manually correlating alerts, logs, and infrastructure data across multiple tools. Traditional monitoring systems generate excessive alert noise without providing actionable insights, leaving on-call teams to piece together context under time pressure and increasing mean time to resolution.
Solution
Aurora by Arvo AI provides an open-source, agentic incident management platform that autonomously investigates cloud incidents and delivers structured root cause analysis. When a monitoring tool fires an alert, Aurora's LangGraph-orchestrated LLM agents dynamically select from 30+ tools to query infrastructure, execute CLI commands in sandboxed pods, search knowledge bases for similar past incidents, and traverse dependency graphs. The agents synthesize findings into a structured root cause analysis with remediation recommendations, reducing the manual toil of incident investigation. Aurora is available as a fully hosted service or can be self-hosted in a customer's own VPC, with all data remaining under the user's control.
Target Audience
Primary customers are Site Reliability Engineers, DevOps teams, and platform engineering groups at cloud-native organizations that need faster incident response and automated root cause analysis across multi-cloud and Kubernetes environments.
Features
- LangGraph-orchestrated multi-agent system that autonomously selects and executes investigation workflows across 30+ infrastructure tools
- Sandboxed pod execution for running CLI commands safely during incident investigation
- Integration with 22+ tools including PagerDuty, Datadog, Grafana, Slack, GitHub, Confluence, Netdata, Dynatrace, Coroot, ThousandEyes, and BigPanda
- Cross-cloud support for AWS, Azure, GCP, OVH, Scaleway, and Kubernetes environments
- Knowledge base search that references similar past incidents to accelerate diagnosis
- Infrastructure dependency graph traversal to identify blast radius and affected services
- Apache 2.0 open-source licensing with free self-hosting and no per-seat or per-incident pricing
- Kubernetes deployment via Helm chart with one-command setup through make init and make prod-prebuilt