MCP Server

What is MCP Server?

MCP Server is a FastAPI-based server that integrates with the EAIM Technologies API, processes natural language queries, and provides data visualization with real-time WebSocket updates.

How to use MCP Server?

Install Python 3.9+, clone the repo, install dependencies with pip install -r requirements.txt, and run uvicorn app.main:app --reload. Alternatively, build and run the Docker image. Send natural language queries to the /api/query endpoint or via WebSocket.

Key features of MCP Server

• Processes natural language queries to extract API parameters
• Integrates with the EAIM Technologies API
• WebSocket support for real-time updates
• Provides data visualization dashboards

Use cases of MCP Server

• Query engagement metrics for a company (e.g., "Vodafone Egypt over the last month")
• Retrieve available dashboards and visualizations
• Build real-time monitoring applications using WebSockets
• Prototype natural language interfaces for API data

FAQ from MCP Server

What does MCP Server integrate with?

It integrates with the EAIM Technologies API.

How do I run MCP Server?

You can run it via uvicorn app.main:app --reload or using the provided Docker image.

What transport does it support?

HTTP at endpoints like /api/query and WebSocket at /api/ws.

What are the runtime requirements?

Python 3.9+ and pip; Docker is an alternative.

What API endpoints are available?

/api/query, /api/dashboards, /api/visualizations/{dashboard_id}, and WebSocket at /api/ws.