Docs
Deploy Operate
Operate runs as a single Docker Compose stack. Pick how you want to run it.
What you are deploying
Operate is a self-hosted AI platform for production investigations that runs inside your own infrastructure. It ships as a single Docker Compose stack containing the investigation agents and the adapters that connect to your logs, databases, and code repositories. Once running, it answers production questions from Slack and Microsoft Teams, investigates alerts, reviews performance, security and cost, traces the root cause, and drafts a fix as a patch for an engineer to review and apply.
Every adapter is read-only. Operate can query logs and records and read code, but it cannot write to production, commit, merge, or deploy. An engineer stays in control of what ships.
Choosing where to run it
Operate works on any cloud provider or on your own hardware: anywhere that can run Docker Compose and reach the systems you want it to investigate. The guides on this page cover a general setup and a step-by-step walkthrough for DigitalOcean. The same stack runs on AWS, Google Cloud, Azure, or bare metal with the same configuration.
You also choose the AI model. With a model hosted inside your network, investigation data stays there. With an external AI provider, the context an investigation needs is sent to that provider, depending on your configuration. Separately, Operate sends a usage record to operatex.dev for billing: your App ID, token counts, timings, and model name. It does not send your logs, records, or code.
Before you start
You will need an App ID from your Operate dashboard, a machine with Docker and Docker Compose installed, and read-only credentials for the systems you want Operate to investigate. Most teams start with one log source and one repository, confirm the first investigations look right, and then add more adapters.