glpi-plugin-aiticketanalysis

This plugin enhances GLPI 11's helpdesk functionality by integrating artificial intelligence capabilities directly into the ticketing workflow. It connects to an AnythingLLM workspace, enabling local inference using open-source language models such as Qwen, Llama, or any model available through LM Studio. The core value proposition is that all processing occurs on-premise — no cloud services are required, ensuring complete data privacy and control over your support infrastructure.

The plugin introduces several intelligent features: automatic ticket analysis that understands user intent from subject lines and descriptions; OCR-powered attachment extraction for images, PDFs, and scanned documents; RAG (Retrieval-Augmented Generation) enabled knowledge base matching that retrieves relevant past solutions; and intelligent suggestion engines that draft context-aware responses based on historical tickets and associated articles. Additionally, optional read-only GLPI MCP integration allows the plugin to query ticket history and user information without modifying any data.

The architecture is designed for scalability: you can choose between CPU-only inference with small models like Qwen 3B or GPU-accelerated processing with larger models such as Qwen 72B. The OCR module supports multiple languages including English, French, German, Spanish, and Italian, making it suitable for multilingual support environments.