Salesforce has launched headless Data 360 for MCP so AI agents can use customer data directly. According to SiliconANGLE on Aug. 19, the service is designed to provide agents with relevant, controlled customer context. It is built to go beyond retrieval and handle data building, transformation, mapping, segment creation and activation.

Data 360 provides more than 200 existing APIs as programmable endpoints. As a result, AI agents can directly call needed Salesforce data and functions without going through a user interface. Teams can continue related work within existing tools and interfaces.

Existing data work required knowing needed fields, schemas, segments and tables in advance, and moving across multiple tools and teams. This sometimes took days to weeks to finish a project. Salesforce explained that Data 360 centralises API access paths so the process can be handled in one go.

The announcement follows headless 360 that Salesforce unveiled in April. While headless 360 connected any system to Salesforce APIs through MCP, the new Data 360 expands the scope to data intelligence, insights, transformation and collection through agents.

Users only need to describe required analysis conditions in natural language. For example, if a user asks to calculate customer lifetime value for electronics purchases while excluding software purchases, the agent writes the query, creates the calculation and then runs it. In the process, it also automatically sets up semantic models, relationships, transformations and formulas.

AI agents such as Claude, Cursor, ChatGPT and Salesforce Agentforce can find and execute needed Salesforce data and functions without pre-specifying exact field names or function calls.

Salesforce directly hosts and manages the Data 360 MCP server. It also provides pre-packaged skills for common tasks such as modelling, mapping, transformation and activation.