Institutional Sponsor

  • Black Twitter Icon

© 2019 by SciPy.

Frank Schlimbach
Oleksandr Pavlyk

Daal4py—Creating a Data Scientist-first Framework with a Hardware-centric Native Library

Designing tools for Data Scientists has been a challenge in the last few years. From the growing needs of the data analytics space to the needs of data scientists, designing tools to meet their needs while simultaneously delivering on performance and features seems like a tough goal. Daal4py is an attempt at making a data scientist-first abstraction to native performance libraries, delivering advanced functionality like distributed machine learning in a one-liner API. In this talk, learn about how this framework was developed, and the method that was used to connect a low-level library while operating well with community packages.