Autopilot overview

Last updated Dec 12, 2024

Context grounding

Context grounding in Autopilot for Everyone allows you to search existing Context grounding indexes to answer user queries.

For example, you want to query the HR documents of your organization, such as time and expense policies, or employee handbooks, using Autopilot. Using Context grounding, an admin would create the index and enable it in Autopilot, which allows Autopilot to search the index to answer HR-related queries.

To use Context grounding, make sure you meet the following criteria:

Relevant PDF, CSV, JSON, DOCX, XLSX, and TXT documents are uploaded in a shared Orchestrator storage bucket.
The index has been already created from the Orchestrator storage bucket.

Managing Context grounding indexes

Index creation is done via the Context grounding tab in Automation Cloud™ > Admin > AI Trust Layer.

Index enablement and management in Autopilot is done via the Context grounding configuration section in Automation Cloud™ > Admin > AI Trust Layer > Autopilot for Everyone.

Once the index is created, enable it in Autopilot through the Context grounding, panel.

Managing access for Context grounding indexes

Access to Context grounding indexes is managed at the folder level. Users can access indexes under the following conditions:

The index exists in the folder
The user is assigned to the same folder where the index is located
The index is enabled in Autopilot

Adding Context grounding indexes

To add a Context grounding index:

Switch to the Context grounding tab.
Select the tenant where you want to add Context grounding indexes.
Select the Add new option. The Add new index panel opens.
In the Index name* field, write a unique name and descriptive name for your index. For example, to create a index for HR documents, one would use "HR" as the index name.
In the Description field, write a description for your index, such as a brief explanation of the content or data to be ingested and queried.
For Data source*, select:
- Storage bucket - if your documents are uploaded in an Orchestrator storage bucket. The storage bucket must assigned to a shared folder. If you select this option, proceed with step 7.
- IS Connection - if your documents are uploaded in a third-party storage system, namely Sharepoint, Google Drive, or DropBox. By establishing an IS connection to these systems, Context grounding can access data directly stored within them. If you use this option, proceed with step 8.
If you selected Storage bucket as a Data source:
1. From the Orchestrator folder* dropdown menu, select the shared folder where the storage bucket with the desired documents resides.
2. From the Storage bucket* dropdown menu, select the storage bucket where the desired documents are uploaded.
3. From the File type dropdown menu, select the file format you want to index. Default value is All.
  We recommend selecting All, in case you have multiple file types you want to ingest. You cannot change this option once the index is created.
4. Select Sync now to add the index to the Context grounding list. The index is added to the indexes list, with the initial status showing Sync in Progress. Once the sync completes, you are notified and the status updates. Select the Refresh icon to refresh the index tables and retrieve their latest status.
  You can also select this option from the menu to the right of the index to ensure the data in the index is up to date with the data source. This only updates the index, not the data source.
If you selected IS Connection as a Data source:
1. From the Orchestrator folder* dropdown menu, select the shared folder where the desired connection resides.
2. From the Select connector* dropdown menu, select the connector where the desired documents are uploaded.
3. Select Add new connection. This option id displayed when no connections of the selected connector exist.
4. From the Select connection dropdown menu, select the connection where the desired documents are uploaded.
5. Select Add new to add more connections for the selected connector.
6. From the Data source location, choose the folder or subfolder where the desired documents are uploaded.
7. Confirm your selection with the Select folder button.
8. From the File type dropdown menu, select the file format you want to index. Default value is All.
9. Select Sync now to add the index to the Context grounding list. This also updates the list of documents in the external repository.

Enabling Context grounding indexes

To enable an index for Autopilot for Everyone:

Select the tenant for which you want to enable Context grounding indexes. Autopilot for Everyone must be installed in the selected tenant.
Expand the Context grounding section.
Select the Enable Index option. The Enable Index In Autopilot panel opens.
From the Index name dropdown menu, select the index you want to add to Autopilot. The list of available indexes in the selected tenant is displayed.
In the Description for Autopilot* field, write a comprehensive description that should help Autopilot understand when to search for the selected index.
For example, for an index value "HR documents", the description could read "HR policy documents on time/expense, paternity, PTO, and employ handbook. Use for questions about company HR policies, providing accurate, up-to-date information to employees."
Select Enable to add the index. The panel closes and you are returned to the Context grounding table.

Important:

Changes are automatically saved.

Editing Context grounding indexes

To edit an index for Autopilot for Everyone:

Select the tenant containing the index you want to update.
For the desired index, select Edit. The Updating index opens.
Modify the desired properties as needed.
Select Save to update the index. The panel closes and you are returned to the Context grounding table.