# Data sources

Connect a data source and Eraser draws architecture diagrams from what you actually run, not from a description of what should be there. There are three steps to it.

## 1. Connect

Open **Settings → [Data Sources](https://app.eraser.io/dashboard/all?settings=dataSources)** and choose what to connect.

<Figure src="/assets/data-sources-picker.png" alt="The Data Sources pane in Eraser, listing the data sources you can connect" border />

Each data source has its own walkthrough:

* [Connecting AWS](/connecting-aws) – an account, or an organization
* [Connecting Azure](/connecting-azure) – a subscription, or a whole tenant
* [Connecting Google Cloud](/connecting-google-cloud) – a project, or an organization or folder
* [Connecting ServiceNow](/connecting-servicenow) – configuration items and their relationships
* [Connecting VMware](/connecting-vmware) – a CSV export, or a live vCenter connection
* [Connecting Cisco](/connecting-cisco) – Meraki Dashboard or Catalyst Center
* [Connecting Ardoq](/connecting-ardoq) – components and references from your workspaces
* [Connecting OrbusInfinity](/connecting-orbusinfinity) – objects and relationships from your tenant

Every connection is read-only, and only team admins can add one. Once connected, it's available to everyone on the team.

## 2. Scan

Eraser reads the data source and builds an inventory of what it holds and how it connects. The first scan runs when you connect; after that, scans run on demand from **Settings → [Data Sources](https://app.eraser.io/dashboard/all?settings=dataSources)**.

## 3. Diagram

Reference an active connection in AI chat and ask for what you want – a VPC, an application service, a site's network, the whole estate. See [Cloud diagrams](/cloud-diagrams).
