Index Workspace Data with Mongo Connector

With the Rocket.Chat AI pipeline server and the Mongo connector deployed, the next step is to connect your Rocket.Chat workspace. In this guide, you'll authenticate to the Mongo Connector, add your Rocket.Chat workspace, and configure which rooms should be indexed into the Rocket.Chat AI pipeline server.

Before you begin, familiarize yourself with these key terms:

• Backfill: A one-time synchronization of a room's existing message history into Rocket.Chat AI pipeline server. It runs the first time a room is queued for indexing.

• Live sync: Continuously synchronizes new messages after the initial backfill completes, keeping the index up to date.

Step 1: Authenticate the connector UI

1. Open your browser and navigate to your Mongo Connector URL.

2. Locate the Session Access section.

3. Paste the Connector API Key you copied during deployment into the Existing API Key field.

Now that the connector is authenticated, you can procced to connect your workspace.

Step 2: Add and configure a workspace

To begin synchronizing data, add your Rocket.Chat workspace to the Mongo Connector.

1. Under Add Workspace, configure the following settings:

   Field	Description
   Workspace Name	Enter a name for your workspace.
   Slug	Enter a unique, URL-friendly identifier (e.g., company-main)
   Mongo
   Mongo URI	Enter the connection string for your Rocket.Chat MongoDB instance. The database must be reachable from the Mongo connector deployment.
   Database	Select the name of the MongoDB database that stores your Rocket.Chat collections.
   Rocket.Chat AI Pipeline
   Pipeline Base URL	Enter the Rocket.Chat AI pipeline server URL.
   API Key	Enter the Master API Key generated during the Rocket.Chat AI pipeline deployment.
   API Secret	Enter the Master Secret generated during the Rocket.Chat AI pipeline deployment.

2. Click Create New Pipeline. The form expands to display additional pipeline configuration fields.

   1. Enter a Pipeline name.

   2. Click Create Pipeline.

      The newly created pipeline is automatically selected, and the default text splitter is configured.

3. Click Save workspace.

Step 3: Index rooms

After saving the workspace, the next step is to choose which rooms should be indexed with the Rocket.Chat AI pipeline server.

1. Navigate to the Room Catalog section. It displays all the rooms in your workspace.

2. Select the rooms you want to index.

3. Go back up to the Selected Rows section and select an action:

   1. Backfill: A one-time synchronization of a room's existing message history into the Rocket.Chat AI pipeline server. It runs once and then stops.

   2. Live + backfill: A one-time backfill of a room's existing message history, followed by continuous live sync to keep the index up to date as new messages arrive, are edited, or deleted.

   3. Dequeue: Stops indexing for a room, cancels any queued work, and removes the room's documents from the pipeline.

4. Click Queue Selected Now.

The Operations Queue section below shows the progress status of each index operation. When the status shows Completed, Rocket.Chat AI pipeline has finished indexing the selected rooms successfully.

For your initial deployment, start by indexing a small room. This lets you verify that your network connectivity, credentials, and permissions are configured correctly before indexing your entire workspace.

Next steps

Your Rocket.Chat workspace is now connected to the Rocket.Chat AI pipeline server, and your indexed data is ready for AI-powered applications. Continue to the AI Search and Rocket.Chat AI app guides to interact with your indexed workspace data.