Straddling the boundary between open-source and proprietary platforms


I’ve written an opinion piece about some work my team are doing with Microsoft and the Alan Turing Institute are doing. The work is exploring how the open-source (Pangeo) communities/software and proprietory services (Azure Machine Learning / AWS Sage Maker) could work well together and really complement each other.

The post is here:

We are really interested in the thoughts of the community so please add your comments here or there.



This is very cool, thank you for writing this up! Using whatever the proprietary cloud platforms offer while also adhering to standard interfaces seems very important going forward.

I’ve written something along this spectrum in Would be very interested in hearing what you think about it!

1 Like

Hi @yuvipanda . Thanks.

Enjoyed your post. I think:

“Use abstractions that allow swappable implementations anytime you have to talk to a cloud provider API directly.”

is what we are trying to achieve in this, but the abstractions are big, high-level ones. I think in some areas we are not clear what those abstractions should be yet…