Wren AI

What it is: Wren AI is an open-source GenBI (Generative BI) agent. Unlike most text-to-SQL tools it ships with a real semantic layer, a chat UI, automatic chart generation, and shareable conversation threads — basically a self-hosted alternative to a hosted "ask your data" SaaS.

Why it matters for data work

Naïve text-to-SQL gets brittle the moment metrics need definitions ("MRR includes annual contracts amortized monthly"). Wren's semantic layer lets you encode those definitions once so the agent answers consistently no matter how the question is phrased.

Install & configure

git clone https://github.com/Canner/WrenAI && cd WrenAI/docker
docker compose -f docker-compose.yaml up -d

Connect a data source through the UI, define your metrics and dimensions in the modeling layer, and start asking questions in the chat panel. Each answer comes with the generated SQL, a chart, and a thread for follow-ups.

Example usage

Replace half of your "quick lookup" Slack pings: a non-technical PM asks "how many active accounts in Europe last week vs the prior week" — Wren answers with a chart and a sharable link, no analyst needed.

Author & links

Author: Canner

Repo: github.com/Canner/WrenAI

License: AGPL-3.0

← Back to Data Agents