You are here:
Create an Individual Retriever
To improve the relevancy of the content returned to a prompt for a particular use case, create a custom individual retriever in AI Models (formerly Einstein Studio). To customize the retriever, select its search index, define filters, and specify what information it returns.
Required Editions
| Available in: All Editions supported by Data 360. See Data 360 edition availability. |
| User Permissions Needed | |
|---|---|
| To manage retrievers or versions in AI Models (create, edit, and delete) | Allow users to manage models in AI Models OR Data Cloud Architect Permission Set |
| To access data with retrievers | For the Knowledge object, make sure that the user has read access to the Knowledge object and has the App Permission: Allow View Knowledge For all other standard objects, make sure that the user has read access to the object and appropriate sharing rules. For more information see, Set Your Internal Organization-Wide Sharing Defaults. |
- In AI Models on the Retrievers tab, click New Retriever.
- Select Individual Retriever as the retriever type. Click Next.
-
Select Data Cloud as the data source for the retriever.
- Select the data space in Data Cloud where the search index resides.
- Select the data model object (DMO) associated with the search index.
- Select the search index.
- Click Next.
-
(Optional) Define filters to narrow the search results.
Retriever filters are available only if the search index that you selected has filter fields defined.
- Select Filter Documents to Return, and specify up to 10 conditions to apply.
-
Select the condition requirements.
For example, to define a specific subset (such as Category = Hardware AND Language = German), select All Conditions Are Met. If different conditions support the same goal (such as Priority = High OR Escalated = True), select Any Condition is Met.
Select Custom Condition is Met to create a custom logic with numbered conditions (such as, 1. Car Make = Honda, 2. Car Model = Civic, 3. CarYear greater than 2000. Create a custom condition like 1 AND (2 OR 3)).
- For each condition, click Add Condition and select a field in the search index, an operator, and a value. The field’s data type, such as text or number, determine which operators are available.
- Click Next.
-
Configure which information to return and how much.
- Specify the maximum number of results to return.
-
Specify one or more fields that ground the prompt. For each field selected, enter a
unique, descriptive label for the retrieved information, and select the field in the
search index.
Select only fields that augment the prompt for your use case and help make prompt responses more accurate and relevant.
- Click Next.
-
(Optional) Turn on citations for this retriever.
- To automatically generate citations that include the source URL, and a heading text, select Standard Citations.
-
To customize the citation, select Custom Citations.
Optionally, enter a base URL, source DMO field that contains the full link or a URL
path that can be combined with the base URL, and the source DMO field to use for the
citation heading.
For example, if the source DMO contains a citation to a specific help page such as Retrieve Data. Enter a base URL of https://help.salesforce.com/s/articleView?id= and the URL path would be the source DMO field that contains the document ID “data.c360_a_ai_retriever.htm”. If the source DMO has a field with the entire URL path, then do not enter a base URL path. Set the URL path to the source DMO field that contains the entire URL path.
Citation headings and links may be surfaced to end-users in downstream applications like Agentforce. Make sure that these fields are populated appropriately and end-users have access to the source links.
- Click Next.
-
Review your selections.
- To change a selection, click the Edit icon.
- Save your changes, and specify a retriever name and an optional description.
-
Save your changes.
The retriever isn’t available for use in prompt templates until you activate a version
When you create a retriever, AI Models creates the first version. Each time you edit and save a custom retriever, AI Models creates another version. Retrievers adhere to the data governance policies set up for your data. Make sure that users who call the retriever have permissions to access the data and knowledge queried by the retriever.

