Managing Building Automation with AI: Claude + Qbee + Domoticz (Part 2)

Don't have time to watch the whole video? You can check out some summary materials in the following notebook instead: https://notebook.google.com/notebook/... Moving beyond gateway-level infrastructure, the real test of distributed facility management is interacting with individual edge devices—sensors, relays, switches, and thermostats—without getting bogged down in fragmented automation consoles. In Part 2 of our AI-native building automation series, we step down into the device layer. By extending our Model Context Protocol (MCP) tooling into Domoticz, we give Claude direct visibility and control over physical building peripherals. Instead of clicking through room menus or manually formatting REST API payloads, you can use conversational AI to query live telemetry, diagnose device states, and safely trigger physical actuators across distributed properties directly from the terminal. In this video, we cover: Moving from Fleet to Device Layer: Bridging Claude Code directly to edge Domoticz instances via MCP Building the Device MCP Server: Exposing device lists, status inspection, and control endpoints as native LLM tools Natural Language Telemetry Queries: Pulling environmental data, battery statuses, and sensor states across multiple zones in seconds Edge Actuation & Control: Toggling switches, managing lighting, and issuing commands directly through plain-English requests Secure Remote Communication: Routing local device commands reliably across remote edge networks What’s Coming Next: Developing multi-property autonomous routines, scheduled workflows, and self-healing automation logic Key Resources & Tech Stack: Qbee Device Management & Remote Access: https://www.qbee.io Domoticz Home & Building Automation: https://www.domoticz.com Claude Code: https://claude.com/product/claude-code Model Context Protocol: https://modelcontextprotocol.io Project GitHub Repository: https://github.com/coevin/domoticz-mc... Subscribe to follow along as we connect complex automation routines, test multi-property workflows, and build a complete AI-driven facility management stack.