WrenAI: Conversational Data Analytics AI Assistant with Direct Access to Answers, SQL Queries & Analytics Reports

General Introduction

WrenAI is an open source SQL AI assistant designed specifically to help data teams, product teams and business teams gain data insights through natural language conversations. It is capable of converting natural language into SQL queries, generating charts, spreadsheets and reports, and supporting multilingual interactions. The project is developed and maintained by Canner under the AGPL-3.0 open source protocol and has received more than 2,800 star ratings on GitHub.The core strength of WrenAI is its complete end-to-end solution that includes an intuitive user interface, a powerful AI service layer, and a semantic engine that securely and accurately handles the need to query data without writing code to obtain data analysis results without writing code.

WrenAI features visual management of data modeling, tagging of business relationships across models, and table tagging of business descriptions for each table and field. The labeled business descriptions are used as the context of the larger model to improve the accuracy of natural language query SQL.

Wren AI Text-to-SQL Agentic Architecture

WrenAI Manage data modeling, view table relationships, note the AI modeling feature in the upper right corner

WrenAI Intelligent Generation of Answers, Queries, Data Reports

Function List

• Multi-language natural dialog: supports dialogic interaction with data in multiple languages
• Intelligent Data Exploration: AI-Driven Data Understanding and Problem Recommendation
• Semantic indexing system: semantic understanding through well-designed UI/UX
• Contextual SQL generation: Combining metadata, schema and terminology to generate accurate SQL queries
• No-code data analytics: data insights are available through conversations
• AI-driven visualization: automatic generation of data summaries and visualization charts
• Data export integration: support for export to Excel and other analytical tools
• Security Assurance: Adopts RAG architecture, no need to expose data to LLM models

Using Help

1. System deployment

WrenAI offers a variety of deployment options:

1. Docker Deployment:
   • Direct deployment using the provided Docker configuration file
   • Ideal for quick startup on a single machine
2. Kubernetes Deployment:
   • Deployment with Kustomization
   • Need to ensure that the following dependencies are met:
     • nginx.ingress
     • external-dns
     • cert-manager
     • kubectl kustomize
     • helm (required for minikube environment)

2. Quick start

1. Basic Configuration
   • Obtain the necessary API key (OpenAI API key)
   • Configuring a Database Connection (PostgreSQL)
   • Setting environment variables and keys
2. data access
   • Connecting your data sources
   • Defining Data Relationships
   • Setting up business terminology mapping
3. Usage Process
   a) Initiate a dialog:
   • Select the relevant data table
   • Ask questions in natural language
   • AI automatically generates suggestions for relevant issues

   b) Data exploration:

   • View Data Structures
   • Understanding field meanings
   • Explore data relationships

   c) Analysis and visualization:

   • Getting SQL Query Results
   • View a summary of AI-generated data
   • Automatic generation of visualization charts
   • Exporting analysis results

3. Advanced functions

1. Semantic modeling:
   • Use of the "Modeling Definition Language"
   • Setting up data relationships
   • Define the calculation logic
2. Data integration:
   • Excel Add-In Integration
   • Data export function
   • Interfacing with other analytical tools

4. Security statement

• Ensure data security with RAG architecture
• No need to expose raw data to LLM models
• Support for private deployment