Telegram AI Chatbot with InfraNodus GraphRAG Knowledge Base

This n8n workflow builds an intelligent Telegram chatbot powered by an AI agent that leverages InfraNodus GraphRAG knowledge bases as 'experts'. Unlike standard RAG vector stores, GraphRAG provides a holistic view of data through knowledge graphs, enabling better relation retrieval and higher-quality responses without complex setup.

The process starts with a Telegram trigger capturing user queries. The central AI agent node evaluates available tools—each linked to an InfraNodus knowledge graph with auto-generated descriptions—and selects the most relevant one. It may refine the query for optimal results before forwarding it via HTTP to the chosen InfraNodus endpoint.

Responses from GraphRAG incorporate full context, structural insights, and relations between data points, making answers more accurate and insightful. Benefits include quick setup (no vector store management), easy updates, and scalability across multiple knowledge bases. Ideal for customer support, research assistants, or domain-specific Q&A bots.

Use cases: Internal knowledge bots for teams, AI consultants drawing from diverse expertise graphs, or enhanced Telegram interfaces for SaaS products needing relational knowledge retrieval.