Semantic models

Semantic models in Flow make your business data available to AI agents securely and reliably via the Model Context Protocol (MCP).

When you create a semantic model in Flow, you automatically get an MCP server with all the tools an AI agent needs to query your business data, analyze it, and build dashboards and reports. This works with native Hypergene AI agents, and with agents built in third-party apps such as Claude Cowork or M365 Copilot.

Typical use cases include:

Building PowerPoint board reports with M365 Copilot or Claude.
Creating operational reports and dashboards.
AI-powered ad-hoc data analysis.
Conversational Q&A over your business data, where users get answers in plain language without writing SQL queries.

flowchart TD
    MCPClient["Any MCP client (AI agent)"]
    Copilot["M365 Copilot (Excel, PowerPoint)"]
    HAI["Hypergene AI"]
    Claude["Claude"]
    MCPServer["Hypergene Flow MCP Server for data analytics"]
    SM["Semantic models 1..n<br/>(Datasets, dimensions, metrics, relationships)"]        
    Data["SQL Server 1..n"]    

    MCPClient --> MCPServer
    Copilot --> MCPServer
    HAI --> MCPServer
    Claude --> MCPServer
    MCPServer <--> SM
    SM --> Data

Table of Contents

Semantic models