The is an open-source Python package that bridges AI assistants — such as Claude — with the Eulerian Marketing Platform APIs. By implementing the Model Context Protocol (MCP) , it allows AI models to directly interact with Eulerian's analytics, attribution, and campaign management data through natural language.
Eulerian is a leading trusted by advertisers and agencies for:
- — server-side, first-party data collection with CNIL exemption and GDPR compliance
- — from Single Touch to Data-Driven and Augmented Data-Driven models across all online and offline media
- — real-time audience segmentation and activation across 100+ connectors
- — AI-powered insights on acquisition, conversion, retention, and lifetime value
With the MCP server, all of this data becomes directly accessible to AI assistants — no manual exports, no context-switching between dashboards.
Once connected, your AI assistant can query Eulerian's APIs to answer questions like:
- "How did our paid search campaigns perform last week?"
- "Which media channels contributed the most to conversions this month?"
- "What is the customer journey breakdown for our top-converting segment?"
This enables , allowing marketing teams to get insights on demand without requiring SQL expertise or deep platform knowledge.
The MCP server is available as a Python package and can be installed via pip:
pip install eulerian-marketing-platform
You will need an active Eulerian account ( type) and API credentials ( type) to connect the server to your platform instance.
Ask questions about campaign performance in plain language
Get attribution insights without opening the platform
Automate reporting pipelines using AI-native tooling
Extend and customize the server to fit internal workflows
Even though our Eulerian Analytics Platform MCP has been designed with security and data quality as top priorities, there are important considerations when using it with AI systems.
Always verify with your IT/Security department which LLM platforms are approved for business data use. Organizations may have strict data processing location policies.
Use enterprise AI services (Claude Teams/Enterprise, ChatGPT Enterprise) or company-managed self-hosted solutions.
LLMs can formulate incorrect queries to the MCP, leading to accurate data for the wrong question.
Always verify the data returned, especially for critical business decisions. In Claude, you can click on any data point to see which properties and metrics were actually queried.
Be specific in your questions, cross-check important insights with the Eulerian Analytics interface, and use MCP for exploration rather than as your sole source of truth.
When using multiple MCPs simultaneously, a compromised or insecure MCP can expose data from all connected MCPs. The LLM may retrieve data from secure sources and inadvertently share it with less secure MCPs.
Pay attention when using MCP and use as many time as possible official sources.
This section is credited to Florian Rieupet - Piano Analytics