Summer isn’t quite over, but the year-end is fast approaching. With that comes end-of-year sales crunch and pulling reports for planning. We’ve been hard at work adding functionality that will make that whole process a breeze.

📄 Crunch through any data format

Regularly work with Excel files or JSON? We bet you do! Now you can uploads those formats to Fabi.ai and slice and dice the data using all our incredible tools (more on this here).

Bonus: If you’re dealing with large datasets you should consider converting you CSV files to Parquet, which we also support.

🏎 Speed & accuracy

If we’re already offering the fastest, most robust data analysis experience, why would we keep investing more here? We believe that every second you have to wait for code to run is a micro-interruption in your analysis and breaks your flow. There’s always more work to do on this front, but we’ve taken massive leaps forward this past month and we know you’ll enjoy it.

By the way, if performance is of the essence for you, here are two tips:

1. Consider using polars instead of pandas

2. Ask the AI to optimize your code - in one click you could literally experience a 10-12X improvement in speed

⭐️ New navigation & favorite Smartbooks

Your Smartbooks and Smart Reports are now organized and you can “favorite” them from the navigation.

🎁 Other goodies

• AI preview: In the AI interface, you can directly preview the code output before adding it to your Smartbook

• Tag DataFrames: @ any DataFrame that you want the AI to target specifically

• Dynamic pick list filters: If you want to turn a DataFrame field into a list of distinct values in a filter dropdown, you can now do that in just a few clicks with “Dynamic” pick lists (docs on creating filters)

• Export to Google Sheets: This is fast becoming a favorite and we’ve made a few improvements (learn more)

Things you might find interesting

• If you’re working on churn prediction, consider using survival analysis

• There are more similarities between librarians and data teams than one might think. This posts reframes the way to think about data availabilities within the enterprise.