Wren AI is an open-source Generative Business Intelligence (GenBI) agent designed for data-driven teams to interact with databases using natural language. It enables users to generate SQL queries, charts, and AI-driven insights from their data, simplifying complex analysis and decision-making.

How It Works

Wren AI employs a Retrieval-Augmented Generation (RAG) architecture. It utilizes a semantic engine to build a contextual understanding of business terms and data schemas, mapping them to data sources. This context, stored in a vector database, is retrieved to guide Large Language Models (LLMs) in generating accurate SQL queries and insights without exposing raw data to the LLM.

Quick Start & Requirements

• Install: Setup is described as achievable within 3 minutes. Specific installation commands are not detailed in the README, but an Installation Guide is linked.
• Prerequisites: Supports various LLMs including OpenAI, Azure OpenAI, Gemini, Anthropic, Groq, and Ollama. Performance is highly dependent on the chosen LLM.
• Links: Installation Guide, Usage Guides, Documentation, API Documentation.

Highlighted Details

• Supports multiple natural languages for querying.
• Features "GenBI Insights" for AI-generated summaries and reports from query results.
• Includes AI-powered data exploration with recommended follow-up questions.
• Offers an API for embedding GenBI capabilities into other applications.

Maintenance & Community

• Community channels include Discord and GitHub Issues.
• A public roadmap is available for tracking upcoming features.
• Discord Server

Licensing & Compatibility

• The README does not explicitly state the license.

Limitations & Caveats

The project's performance is heavily reliant on the quality of the selected LLM, with less capable models potentially leading to reduced accuracy and slower response times.