SQLBot is an intelligent data querying system leveraging large language models (LLMs) and Retrieval-Augmented Generation (RAG) to enable high-quality text-to-SQL conversion. It targets users and applications needing to interact with databases using natural language, offering ease of integration and robust security features.

How It Works

SQLBot combines LLMs with RAG to translate natural language questions into SQL queries. This approach aims to improve accuracy and relevance by grounding the LLM's responses in specific data context retrieved via RAG. The system is designed for out-of-the-box usability and seamless integration into existing workflows and AI platforms.

Quick Start & Requirements

• Installation: Requires a Linux server with Docker and Docker Compose installed.

  mkdir -p /opt/sqlbot
  cd /opt/sqlbot
  curl -o docker-compose.yaml https://raw.githubusercontent.com/dataease/SQLBot/main/docker-compose.yaml
  docker compose up -d
  

• Access: http://<your-server-ip>:8000/ with default credentials admin/SQLBot@123456.
• Deployment: Also available via the 1Panel application store.

Highlighted Details

• Open-source BI tool DataEase, server management panel 1Panel, and enterprise AI platform MaxKB are sister projects.
• Supports integration with AI application development platforms like n8n, MaxKB, Dify, and Coze.
• Features workspace-based resource isolation and fine-grained data permission control for security.

Maintenance & Community

• Project is part of the Feizhiyun ecosystem.
• Technical discussion group available for support.

Licensing & Compatibility

• Licensed under FIT2CLOUD Open Source License, which is based on GPLv3 with additional restrictions.
• Compatibility for commercial use or linking with closed-source projects requires careful review of the license terms.

Limitations & Caveats

The FIT2CLOUD Open Source License's additional restrictions may impact commercial use or integration into proprietary systems, necessitating a thorough understanding of its terms.