identity-rag-customer-insights-chatbot
Basic Information
This repository provides a demonstrator application for IdentityRAG, a retrieval-augmented generation system that integrates identity resolution to give context-aware answers about specific customers. It unifies customer records from multiple, disparate sources, builds a deduplicated golden record for each customer, and then uses that unified view as grounding for responses generated by any supported large language model. The project includes a conversational demo built with Chainlit and a Streamlit live demo, and it shows how to connect to a Tilores customer data instance. The repo includes instructions to install dependencies, configure environment variables for Tilores and an LLM provider, and run a local chat server to query customer data and retrieve actionable insights.