Dreamdata launches Dreamdata AI with an Analytics Agent and MCP server

Dreamdata announced Dreamdata AI on August 31, 2026, a set of AI capabilities built on its account-based B2B data model. According to the company blog, the release has three parts: an Analytics Agent inside Dreamdata that builds and interprets reports from plain-English questions, an MCP server that brings the same functionality into LLMs a team already uses, and support for custom agents running on Dreamdata's exported data warehouse. Dreamdata says all three run on the same data model and a governed semantic layer that fixes definitions such as pipeline, revenue and attribution, and that answers can be traced to the metrics, filters and attribution models used.

The details

• The Analytics Agent and MCP server are included on all Dreamdata paid plans at no added cost.
• Custom agents run on the data warehouse add-on, where customers export Dreamdata's account-based data model.
• A report configurator shows the metrics, filters, date range, segmentation and attribution model behind each answer.
• Prospective customers can try the Analytics Agent or MCP server with their own data on a free trial.
• Dreamdata names Claude, ChatGPT and Gemini as LLMs B2B marketing teams use for analytics work.

Context

This is the second Dreamdata AI connectivity release in our coverage, following its February 18, 2026 work connecting its data warehouse to AI agents through MCP. It also follows the Analytics Hub for custom reports and dashboards launched on June 2, 2026. A companion post describes the data foundation behind the release.

Not disclosed

• Price of the data warehouse add-on
• Usage limits for the Analytics Agent or MCP server
• Named customers