Last week, our core engineer Jimmy at Wren AI was honored to be invited as a speaker at the Apache Airflow & dbt Taipei Meetup to share about “Why Text-to-SQL is so hard and how Wren AI aims to solve it” and what we learned while developing Wren AI. It was a fantastic night, meeting with many amazing developers to discuss how Wren AI can help companies and data engineers/analysts gain more insights with less effort.

The event was incredibly thrilling, with over 100 enthusiastic participants joining us online and in person. We had an amazing time discussing how Wren AI will revolutionize the future of text-to-SQL.

In this sharing, we cover topics about:

1. Historical view — before and the advent of LLM
2. Status quo — after the advent of LLM
3. Why Text-to-SQL is so hard?
4. Introduction to Wren AI
5. How Wren AI aims to solve it?

You can check the full video on YouTube (This presentation is in Mandarin).

The whole slide is also available here (In English): https://hubs.ly/Q02GkQ9q0

What is Wren AI?

Wren AI is an open-source text-to-SQL solution for data teams to get results and insights faster by asking business questions without writing SQL. Wren AI supports a wide range of data sources, such as DuckDB, MySQL, Microsoft SQL Server, BigQuery, etc… Also, she supports open and local LLM inference endpoints such as OpenAI GPT-3-turbo, GPT-4o, and local LLM hosts via Ollama.

If you are interested in Wren AI please check out the following links.

👉 GitHub: https://github.com/Canner/WrenAI

👉 Discord: https://discord.com/invite/5DvshJqG8Z

👉 X: https://twitter.com/getwrenai

👉 Medium: https://getwren.ai/blog

Don’t forget to give ⭐ Wren AI a star on Github ⭐ if you’ve enjoyed this article, and as always, thank you for reading.

Some event pictures of the night. Thank you, everyone, for joining us tonight, see you next time!

‍

‍