Voice is pretty expensive unfortunately. We pay around $10 cents per minute for the voice interaction. LLM calls not included.

Can you just use the browser text-to-speech API?

Yes, we're using OpenAI embeddings

- Mathis from ChatPDF

Why shouldn't they? We haven't been in contact with anyone from Google yet but I expect they are happy about tooling around their products!

Thanks! That's a good idea, we'll put info about the numbers of read/write/delete operations that Firefoo uses up somewhere. For long-running tasks (export, delete/rename colls) we show those numbers in the Tasks popup already but it'd be nice to show them up front!

Hi HN! About a year ago we started using Cloud Firestore at work. And while I love it for its amazing realtime and effortless scaling capabilities, I grew more and more frustrated with the limitation in tooling around it. No way to query from shell, export collections, rename docs, edit nested field names, …

One day I had enough and pitched an idea to a friend: to create the experience I wanted to see. Luckily he was all in and we continued to spend our Corona lockdown hacking together an Electron app in TypeScript that makes my work delightful again! A few months later and we're glad to announce that Firefoo is now live for everyone to use!

Some features include

- Table View

- Export to JSON and CSV

- JavaScript Query Shell

- Batch Edit Documents

- Geo Point Map

We'd love to hear your thoughts and will be around in the comments all day!

- Mathis & Moritz

Looks great! How did you make the decision to go Electron vs. a web app?

Thanks! It's for security reasons. We want to keep all the data locally on the user's computer, so your data will never touch our or a third party's server (other than Google). If we piped the requests through a backend, nobody could verify that we're not reading the content of your data or secrets.