Show HN: Loofi – Our AI-Powered SQL Query Builder | Hacker News Reader

Show HN: Loofi – Our AI-Powered SQL Query Builder

loofi.dev

19 pointsdaledoback7 comments

Hi HN Community,

We're a team of software engineers who've experienced the complexity of crafting SQL queries. To address this, we've developed a tool that simplifies this process.

Introducing Loofi, an AI-powered SQL Query Builder. Loofi streamlines query building and leverages AI algorithms for intelligent recommendations.

We value community feedback and invite you to try out Loofi. Your thoughts and suggestions will help us improve our tool.

Here's the link to Loofi: https://loofi.dev/

For any questions or assistance, feel free to reach out. We're eager to engage with this community.

Your valuable feedback is much appreciated!

Open articleView on HN

Hello, cool demo. Can you share more about the tech behind this? What’s your objective? Will you release an OSS product?or a SaaS?

Thank you for checking it out! We are running a Python Flask API, and GPT-3.5 as our model. The objective is a SaaS Product, but we really want to figure out who this product can help the most. Any suggestions would be really appreciated.

I run projectbaseball.org, a hosted data warehouse for baseball fans to query a bunch of different data sources. English to SQL would really be a game changer in terms of who can use it. Sql is still a huge barrier.

Things like “which pitcher is throwing more than 2 mph harder this year compared to last year” or “who has changed their mix of pitches”. Fantasy players use these kinds of queries to spot players who have made adjustments and thus shown true improvement (versus the noise inherent in the game itself)

(Note, backend is down while I move from a dedicated server to google cloud)

Cool website, I think that totally would be a game changer for your visitors to more easily query data! Great idea

There's a company/open source community where they automatically parse government databases. Makes it as easy as a git clone (obviously using different syntax) for you to have a local copy - think things like holidays, weather histories, open government data (in the US).

I think journalists would love to be able to query that data via natural language. Things like "what is the most common zip code for a specific traffic violation over the past 3 years" would be super useful.

I can't remember the name but will do some searching later.

https://www.splitgraph.com/

Here is a LLM database project also now in ShowHN you might compare what you have under the hood with:

https://news.ycombinator.com/item?id=36038815

Reply on news.ycombinator.com