Skip to content

Getting started

Connect a source, ingest, and search — in that order.

What you need

An AnvaiOps account (sign up at anvaiops.com) and admin access to at least one source system (Jira, Slack, Confluence, …).

1. Create an account

Sign up and pick your cloud at anvaiops.com. Each account is bound to one cloud at signup — your data stays in that cloud's region.

2. Connect a source

From the console, add a connector and authorize it. Connectors are read-only — they observe your source, never write to it. See Integrations for per-source setup and the scopes each one needs.

Authorize the OAuth app, choose the projects to index. AnvaiOps pulls issues, comments, and resolutions.

Install the app to the channels you want indexed. Threads and their resolutions become searchable memory.

Authorize and pick spaces. Runbooks and postmortems become retrievable context.

3. Ingest

Ingestion runs as a scheduled job per source. The first run backfills history; later runs pick up changes from a watermark. You pay for ingest once — queries reuse it.

Ask a question in the console, via the API, or through an agent over MCP. Results are tenant-scoped and ranked by hybrid retrieval (vector + keyword) over your corpus.

curl -s https://<your-workspace>.anvaiops.com/api/search \
  -H "Authorization: Bearer $ANVAIOPS_API_KEY" \
  -d '{"query": "payments webhook 500s after deploy", "top_k": 5}'

Next