Canner/WrenAI

GenBI (Generative BI) for AI agents, an open-source, governed text-to-SQL through an open context layer that turns natural-language questions into trusted dashboards, charts, and SQL across 20+ data sources, such as BigQuery, Snowflake, PostgreSQL, ClickHouse, Amazon Redshift, Databricks and more.

What it solves

WrenAI addresses the problem of AI agents producing untrustworthy or incorrect SQL queries when interacting with business databases. It solves this by providing a governed semantic layer and an AI context layer that ensures agents use approved business definitions, metrics, and company knowledge rather than guessing based on raw database schemas.

How it works

WrenAI acts as an engine that sits on top of existing data warehouses and transformation pipelines. It uses the Modeling Definition Language (MDL) to create a reviewable, version-controlled semantic layer of models, cubes, and metrics. This is paired with an AI context layer—consisting of memory, examples, and unstructured knowledge (like wikis)—to guide agents. The system employs dry-plan validation and structured errors to ensure generated SQL is correct and governed before execution across 22+ supported data sources.

Who it’s for

It is designed for agent builders and data teams who want AI agents to produce trustworthy business intelligence (BI), including governed text-to-SQL answers and shareable dashboards, without locking business logic into a specific vendor's UI.

Highlights

  • Governed Text-to-SQL: Uses a semantic layer to ensure queries are based on business truth, not just schema guesses.
  • Agent-Driven Workflow: Provides a CLI and discovery stubs that allow AI agents (like Claude Code or Cursor) to set up projects, enrich context, and execute queries.
  • GenBI Dashboards: Enables agents to generate and deploy browser-side dashboards to platforms like Vercel or Cloudflare Pages.
  • Open and Version-Controlled: The semantic layer (MDL) and context are Git-friendly and reviewable.
  • Broad Connectivity: Supports 22+ data sources via an Apache DataFusion-based engine, including BigQuery, Snowflake, PostgreSQL, and DuckDB.

Related

  • Project
    xpert-ai/xpert

    An open-source Agent platform that uses a hybrid agent-workflow architecture and a unified plugin system to build governed, enterprise-grade agentic applications.

  • Project
    lightdash/lightdash

    An open-source agentic BI platform that uses a governed context layer to power dashboards, AI agents, and data apps with a software-engineering workflow.

  • Project
    tencentmusic/supersonic

    An AI-powered BI platform that combines a natural language chat interface with a headless semantic layer to provide reliable, governed Text2SQL data analytics.

  • Dispatch
    Inside OpenAI's In-House Data Agent

    OpenAI has developed a bespoke internal AI data agent that enables employees to perform complex data analysis across 600 petabytes of data using natural language, utilizing a multi-layered context system and a self-learning reasoning loop.

  • Project
    open-metadata/OpenMetadata

    An open-source context layer and metadata management platform that provides AI agents and users with the governed business semantics, lineage, and organizational memory needed to trust and use enterprise data.