--- schema_version: "1" id: 46e871dbe00614190072448488854fb02180cf849a71a419b318c1da0e708d8b title: Test an Ensemble Retriever collection: admin version: current locale: en-us portal_url: https://docs.salesforce.com/admin/current/en-us/data/264-0-0/c360_a_ai_retriever_test_ensemble.md canonical_url: https://help.salesforce.com/s/articleView?id=data.c360_a_ai_retriever_test_ensemble.htm&release=264.0.0&type=5 source_site: help.salesforce.com content_type: documentation content_hash: 503a1f4ea74439a4012fb9928573ef1454e98dd4d5ece7df54735e85b74d45aa description: Test an ensemble retriever configuration in the AI Models retriever playground before you use it in prompt templates or AI agents. Add or remove retrievers to an ensemble, configure test data, and result parameters to verify the retrieval logic. products: - Data 360 release: "264" --- # Test an Ensemble Retriever Test an ensemble retriever configuration in the AI Models retriever playground before you use it in prompt templates or AI agents. Add or remove retrievers to an ensemble, configure test data, and result parameters to verify the retrieval logic. ### 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, test, and delete)

Allow users to manage models in AI Models AND Access to Field Relationship Path object

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 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.
#### See Also - [Create an Ensemble Retriever](https://help.salesforce.com/s/articleView?id=data.c360_a_ai_retriever_ensemble_create.htm&release=264.0.0&type=5 "Use ensemble retrievers to get the most relevant information from multiple searches or data sources. In AI Models (formerly Einstein Studio), an ensemble retriever is a collection of individual retrievers. When you run an ensemble retriever, it runs all the individual retrievers, combines their results into a single list, reranks the list according to relevance to the search request, and returns just the most relevant information to the prompt or agent.") ## Customize an Ensemble Retriever in the Playground 1. In AI Models on the Retrieve tab, select **Test Retriever**. Alternatively, select **Test Retriever** for an ensemble retriever from the drop-down action menu. Ensemble retrievers that contain Agentforce Data Library (ADL) retrievers can't be tested in the playground. 2. Select a mode for testing the retriever. In the basic mode, the retriever playground returns a source recall score and a response time. In the advanced mode, the playground returns an AI-generated context recall score and a context precision score based on an expected answer to a test question. For more information on metrics and costs, see, [Retriever Metrics and Results](https://help.salesforce.com/s/articleView?id=data.c360_a_ai_retriever_test_metrics.htm&type=5&language=en_US) and [Billing Considerations for Testing Retrievers](https://help.salesforce.com/s/articleView?id=data.c360_a_ai_retriever_test_billing.htm&type=5&language=en_US). 3. Select an ensemble retriever that you want to test and validate. When you search for retrievers, only active retrievers are listed. The retriever configuration panel shows the individual retrievers that are part of the ensemble retriever. 4. To customize or reconfigure the ensemble retriever, select **Edit**. 5. Select or deselect retrievers by data space. An ensemble retriever can contain a maximum of 10 individual retrievers. The **Select Retriever** tab shows the retrievers that are part of the ensemble. 6. Modify the number of results that you want the retriever to return. If the results to a user's question don't include the expected response, increase the number of results returned. 7. Customize the fields returned by the retriever by adding or removing existing fields. ![Note](https://sf-zdocs-cdn-prod.zoominsoftware.com/tdta-data-data_cloud-264-0-0-production-enus/f3569962-3c9b-4f83-9d0c-b6a574e785da/images/icon_note.png) Note The retriever returns these new fields when you save the configuration and use it with agents. 8. (Optional) Turn on **Citations Settings** to preview and validate citations from individual retrievers. ## Test Your Retriever Configure expected source data and enter a natural language query to test your retriever. Expected source data is the desired source from where the correct answer is expected. With ensemble retrievers, the desired source records can be part of different retrievers. In the playground, you can select test data from each individual retriever that is part of the ensemble. 1. To test your retriever responses from expected sources, select **Configure Expected Source Data**. 2. In the **Configure Expected Source Data** window, select a retriever. 3. Select data model object if you know which records contain the response or select keyword search. For hybrid search retrievers with keyword matches, select test data by keyword search or data model object records. Otherwise, you can only select by data model object records. Select data model object records if you know which records contain the response to your test question. Select records based on a keyword search if you know what your response should include, such as, for a question related to paid time-off, you can enter “PTO or Time-off.” 4. Repeat steps two and three for each retriever that is part of the ensemble. You can select up to five test data records for each individual retriever in the ensemble. 5. Ask a test question. In the advanced mode, you can test your retriever responses based on an expected answer to the test question. After you run your test, the Test Retriever panel shows the number of documents retrieved by the retriever. In the Results panel, review the overlap between the top chunks returned by the retriever and the expected source data chunks. Verify validity of sources with links to citations for the retrieved chunks. Customize retriever configuration with your test data to get the optimal response. Save tested configurations as new versions.