Overall Satisfaction with IBM Watson Studio on Cloud Pak for Data
The IBM Watson Studio is mainly used for one single department, the data science team. It mainly addresses the devops overhead of heavy jupyter notebooks and provides an integrated interface for people who are not familiar with infra and storage. It also provides a point of integration with other IBM services.
- Sharing with team
- GitHub integration
- Free pricing plan if you want to try things out
- Loading times can be slow
- Tabs can be hard to navigate
- not enough out of box examples
- Made running experiments more streamlined
- Reduced devops overhead
- Sometimes does mean integration with things that are not on IBM harder
AWS Sagemaker is a well-established product that supports on-demand notebooks, data pipelines, and so on, however, it also comes with the learning overhead of the whole AWS stack. It does allow per-defined models, but the benefit of using IBM Watson Studio is that users are able to leverage per-trained models and significantly reduce training time.