Query your Companion.energy data in plain language, from the AI assistant you already use.

The Analytics MCP is a remote Model Context Protocol server that lets AI assistants — Claude, ChatGPT, Microsoft 365 Copilot, Cursor and any other MCP-capable client — securely read your energy analytics. Ask a question in natural language and the assistant retrieves the answer live from the platform: assets and sensors, consumption and production, costs and revenue, savings, forecasts, and market prices.

It is read-only. The Analytics MCP can look up and analyze your data; it can never change, create, or delete anything on the platform.

What you can do

Point your assistant at the Analytics MCP and ask questions like these. Each one maps to live data — no dashboards, no exports.

Assets & sensors

• "List all the assets for customer Acme, and tell me which ones are batteries."
• "Which of my assets stopped reporting sensor data in the last 24 hours?"
• "What's the state of charge of the battery at the Antwerp site right now?"

Consumption & production

• "Show hourly consumption for the Antwerp plant last week and highlight the peak."
• "Compare solar production against consumption for site X this month."
• "What was the highest peak demand across my portfolio yesterday, and at what time?"

Cost, revenue & savings

• "Break down energy cost by site for Q2 and split it between energy and grid fees."
• "How much revenue did the battery at site X earn last week?"
• "Rank my sites by realised savings over the last 30 days."

Market prices

• "What were the Belgian day-ahead prices yesterday, and when was the cheapest hour?"