> For the complete documentation index, see [llms.txt](https://hi-primus.gitbook.io/bumblebee/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://hi-primus.gitbook.io/bumblebee/undefined.md).

# Hi Bumblebee!

Bumblebee is the easiest, and most powerful tool to clean, transform, and prepare data of any size for Analysis, Visualization, Reporting, and Machine Learning. All in a spreadsheet-like interface. ![Absolute value on the interface](/files/-MY7Y10RkN6gLrg4_56d)

Bumblebee can run ever multiple engines like Pandas, Dask, cuDFD, Dask-cuDF, Spark or Vaex.

The engine you decide to use will depend on the resource you have at hand.

## Some features

This is what Bumblebee can offer you.

### Clicks & Drags-and-Drops

Our spreadsheet like-interface makes the work easier for you to wrangle the data, correct wrong and duplicate values, and identify and group similar strings.

### Automated Workflows

Create cleaning repeatability by creating a “Data Recipe”. Next time you need to clean the datasets from the same source, run your pre-saved Data Recipe, and make Bumblebee clean the data for you.

### Load data from anywhere

Load data from CSV, JSON, Parquet, local, from a URL or from a remote storage system. Also, connect to data from databases and start working on your data wrangling in no time!

## Prerequisites

* OS Support
  * Ubuntu 18.04 LTS/20*\**
  * Windows 10
  * GPU Support (Dask-cuDF, cuDF)
  * Pascal or Better  
  * Compute Capability
* CUDA Support (Dask-cuDF, cuDF)
  * 10.2
* Python Support
  * 3.7
  * 3.8