Our first podcast: Sample Space

There are many libraries, tools and techniques out there in the world of machine learning. These usually start out as a mere idea though. That idea phase is very interesting and it's is something we wanted to talk about some more. We've already been producing general educational content for the scikit-learn ecosystem, but also wanted to have a place where we might share some outside perspectives that are more about ideas than tools.

This is why we've started a podcast called Sample Space. It gives us a place to give some extra attention to people who are experimenting with ideas that more people should know about. This includes toolmakers, but also practitioners from the field. Our goal is to have an episode at least once a month, but we may also have periods with more interviews. 

You can follow the podcast on most podcast players including apple podcasts, spotify and rss.com, but we also host each interview on our YouTube channel. If you go there right now, you should see the first three episodes.

The first episode is with Trevor Mantz, the creator of anywidget. It's a great tool that makes Jupyter notebooks more interactive and it's a tool that we've also used internally to rewrite drawdata.

The second interview was with Phillip Cloud from the Ibis framework. This tool allows you to write DataFrame code that can run on many different backends, including some that can technically only handle SQL like DuckDB.

The third episode is with Leland McInnes, who is most well know for creating UMAP but when you have a look at his Github repository you'll notice he's got a long impressive resume of tools at this point.

We hope that you'll enjoy our content. Do let us know if you feel that there are other people to be taken into consideration!