dataease/SQLBot
🔥 基于大模型和 RAG 的智能问数系统，对话式数据分析神器。Text-to-SQL Generation via LLMs using RAG.
What it solves
SQLBot enables conversational data analysis (ChatBI), allowing users to extract data information and generate visualization charts using natural language instead of writing complex SQL queries manually.
How it works
It combines Large Language Models (LLMs) with Retrieval-Augmented Generation (RAG) to perform high-quality Text-to-SQL conversion. The system can be further refined through custom prompts, terminology libraries, and SQL example calibration to ensure the generated queries match specific business scenarios.
Who it’s for
It is designed for users and organizations that want to implement an intelligent data questioning system without extensive development, as well as developers looking to integrate ChatBI capabilities into existing applications like n8n, Dify, or MaxKB.
Highlights
- Out-of-the-box deployment: Quick setup via Docker with simple LLM and data source configuration.
- Security and Control: Features workspace-level resource isolation and fine-grained data permission settings.
- Flexible Integration: Supports Web embedding, popup embedding, and MCP calls.
- Continuous Improvement: Improves accuracy over time using user interaction data and custom terminology libraries.