​

Logging in & creating an account

To get started:

1. Navigate to https://app.fabi.ai/
2. Click on Sign in with Google and log in with your Gmail account

You can log in with your personal email account, however, most collaboration functionality is only available for corporate accounts.

​

Uploading a file

The quickest way to get started is simply by uploading a file that has some data that you want to analyze. Uploading a file in Fabi.ai stores it in DuckDB, allowing you to analyze large files with none of the performance issues. Fabi.ai currently supports CSV, Parquet, Excel and even JSON! (for more information on file uploads see our Uploading Files page)

From the landing page, click Start analyzing a file.

Alternatively, if you’re already in a Smartbook, select Upload Files in the left hand panel.

The first time you upload a file, Fabi.ai will automatically create a SQL cell with the proper syntax for your. Press Run SQL to view your dataset.

​

Asking AI to analyze your data

Whether you’re viewing data from a file you’ve uploaded or from your data warehouse, the output of all queries will be stored as a pandas dataframe.

To analyze this data, go to the right hand panel and ask the AI to analyze your data.

Some tips on using the AI:

• You can tag elements using @ or mention elements (example: “Use dataframe1 to compute the average team revenue by year”)
• You can ask the AI to use SQL or Python specifically
• Once the AI has generated code, you can choose to Apply or Create new cell

Some example prompts:

• “Please optimize @python_code to run faster”
• “Which region has seen the biggest increase in revenue YoY?”
• “Create a scatter plot showing X vs Y and mark the outliers in red.”

​

Using Python for data analysis

As you’ve seen with the AI, you can conduct your exploratory data analysis using Python by leveraging dataframes produced by SQL cells. If you don’t want to use AI to analyze your data, simply create a new Python cell ****and start coding.