Kangas is an open-source tool provided by Comet, provider of an MLOps platform for machine learning teams. 

According to Comet, the tool is intended to help teams explore, analyze, and visualize large-scale multimedia data. 

Kangas offers users a Python API for logging large tables of data as well as a visual interface for performing complicated queries against the users dataset. 

“A key component of data-centric Machine Learning is being able to understand how your training data impacts model results and where your model predictions are wrong,” said Gideon Mendels, CEO and co-founder of Comet. “Kangas accomplishes both of these goals and dramatically improves the experience for ML practitioners.”

With Kangas, customers gain several benefits such as scalability with Kangas DataGrid, the class for representing datasets and the ability to quickly group, sort, and filter across several data points with a simple UI.

Additionally, this open-source tool brings users interoperability by enabling running in a notebook or as a standalone app, both locally and remotely. 

Finally, this offering brings users integrated computer vision support so they can visualize and filter bounding boxes, labels, and metadata without any additional setup. 

For a live demo of Kangas, click here.