---
schema_version: "1"
id: 828d79062dc701b2c3e9a9c843b7afae31984d29b8339558ff91ae75c3e87c16
title: Fine-Tune, Test, and Optimize Retrievers in Retriever Playground (Generally Available)
collection: admin
version: current
locale: en-us
portal_url: https://docs.salesforce.com/admin/current/en-us/release-notes/260-0-0/rn_cdp_2026_spring_retrieverpg_ga.md
canonical_url: https://help.salesforce.com/s/articleView?id=release-notes.rn_cdp_2026_spring_retrieverpg_ga.htm&release=260&type=5
source_site: help.salesforce.com
content_type: documentation
content_hash: e760aea98d7883fad01eea09210649998f981001a69d913bbc56d416db1b8bf2
description: Customize, test, and validate retriever responses in the Data 360 Einstein Studio Retriever Playground. A misconfigured retriever can lead to irrelevant output, erode user trust, and increase costs. You can verify response accuracy directly in the playground without making calls to an LLM or performing manual testing in local environments.
products:
  - Cross Cloud Packages Solutions
release: "260"
---
# Fine-Tune, Test, and Optimize Retrievers in Retriever Playground (Generally Available)

Customize, test, and validate retriever responses in the Data 360 Einstein Studio Retriever Playground. A misconfigured retriever can lead to irrelevant output, erode user trust, and increase costs. You can verify response accuracy directly in the playground without making calls to an LLM or performing manual testing in local environments.

**Where:** This change applies to Data Cloud Developer, Enterprise, Performance, and Unlimited editions. Not available: Government Cloud.

**When:** This feature is available starting in February 2026 in the AMER region and in March 2026 across all regions.

**Why:** Consider a scenario where an agent answers field service questions on how to install double-door refrigerators. Technicians report that the agent responses are incomplete and include only five steps of an 11-step article. The admin tests the retriever in the playground and finds that the article that contains the response is split between two chunks. The top chunk includes the keyword “double-door” and appears in the response, but the second chunk ranks lower and is excluded. The admin increases the result count in the retriever playground to include the second chunk and validates the fix before deployment.

#### See Also

- [_Salesforce Help_: Customize, Test, and Validate Retrievers with Retriever Playground](https://help.salesforce.com/s/articleView?id=data.c360_a_ai_retriever_test.htm&language=en_US)