Introducing Evidently: Open-Source Tool To Analyze Data Drift

?? Is your machine learning model still relevant?

Once you get to production, that is a valid question to ask. Things change, and the model might not keep up. 

To monitor it, we need to keep an eye on data drift. Basically, our goal to check if your data still looks the way it did before - when you trained it, for example. And if you detect the change, understand where it comes from.

We just released an open-source tool to help with that.

It helps evaluate and analyze data drift by comparing the two datasets.

What is great about it?

?? We integrated statistical tests, so you don’t have to.

?? Interactive plots to explore each feature.

?? You can export it as .html to share it around.

?? Works right in your Jupyter notebook.

And yes, it is open-source!

Check out our launch blog post, and try it out on Github.

It is the first step in our vision of creating open-source tools to analyze, debug and monitor machine learning models. Stay tuned for more, and connect here on Linkedin, on Github, or via [email protected].