4 Data 360 Lessons for Connecting Product Knowledge to AI
Salesforce, Wednesday, November 4th, 2026
Salesforce shares four lessons from grounding its Help Agent with Data 360, including assessing content quality before indexing.
Salesforce uses Data 360 to organize product knowledge and Agentforce to power its Help Agent, which answers customers using retrieval augmented generation.
Its lessons are to assess content before building, treat each agent change as a hypothesis to test, define good answers up front with a golden data set and acceptance criteria, and choose testing methods that build confidence.
Salesforce scores articles with signals such as Self-Help Failure Rate and Problem Resolution Rate and pairs them with top case drivers to find gaps.
It built a separate agent for the Informatica help portal, which now resolves 92% of questions without a case.