SRE Platform is an AI incident investigation platform built to reduce investigation time and mean time to recovery (MTTR). It gathers context from your connected tools, weighs likely causes, and recommends the next step in your incident conversation.

Get started · Connect your tools · Documentation

Full-page captures from the demo workspace. Select an image to view it at full resolution.

Spend less time rebuilding context

• Start with an investigation. When an alert warrants investigation, the platform opens an incident and gathers evidence from your connected tools.
• Get to the next decision. Review likely causes, supporting evidence, and a recommended next step together. Missing evidence and uncertainty stay visible.
• Continue from what is known. Ask follow-up questions in Slack or the dashboard. The investigation builds on stored evidence instead of starting over.

See how triage works and the incident workspace for the investigation flow and its limits.

See the investigation context

Know what needs attention

The operational dashboard brings together incidents needing a person, nearby deployments, and gaps in diagnostic coverage.

Find what changed

The changes feed brings code, review, CI, and release activity into one timeline, with filters to narrow the investigation.

Local development

Requires Bun and Docker, plus Node.js for the test suite. See the prerequisites for required versions.

cp .env.example .env        # then set SECRETS_MASTER_KEY (openssl rand -base64 32)
bun install
bun run dev:setup           # start local services, wait for health, apply migrations
bun run dev

The dashboard is then on http://localhost:45173 and the API on http://localhost:43000. Run the repository checks with:

bun run typecheck && bun run test

Local development covers the rest: passwordless dev sign-in, the optional Kubernetes port-forward tunnels, and bun run db:reset.

Explore the project

• User and operator guide
• Workspace layout
• Domain glossary
• Contributing