Great services for fast and effecient data analytics!
January 04, 2021

Great services for fast and effecient data analytics!

Anonymous | TrustRadius Reviewer
Score 8 out of 10
Vetted Review
Verified User
Review Source

Overall Satisfaction with IBM Watson Studio (formerly IBM Data Science Experience)

This system is currently being used with a few students on a data science degree within the School of Computing at our University. We are using IBM Watson as a means to overcome the hardware limitations we have within the our work setting. IBM Watson provides student with access to high powered machines allowing them to run complex machine learning algorithms without having to worry about hardware negatively effecting the performance of said algorithms. It is also a relatively simple system to use, making it a useful teaching tool which requires minimal support for academics. Students have provided positive feedback regarding the use of this service and we plan to expand our use of Watson Studio throughout our other degree options.
  • Clear distinction between services provided.
  • Jack of all trades without being a master of none.
  • Complex processing without an major latency.
  • Some aspects of the UI can be overwhelming for a novice user.
  • Integration with some non-Watson Studio services is limited.
  • It has helped streamline our Data Science courses and provided students with an alternative to needing to have a high end machine at home.
  • It has been especially useful for maintaining our quality of service objectives during Covid as we have maintained and met expectations by using IBM Watson Studio as a key replacement service.
I believe these time savings have been achieved however, this has primarily been for students with a novice level of knowledge around data science and the intricate processes included in the cataloging, refiners, visualization etc. stages of a project. Members of staff with a much broader knowledge of these processes and services have found that a reduction of wasted time can be found on occasion but because they have been trained to use other tools, they find they work more effectively with those.
Some members of staff were upskilling using courses that utilised IBM Watson Studio and they found it to be a particularly useful tool for accessing a large variety of kernels, data sources and compute integration. Having access to a community of 20 million data scientists and the ability to collaborate with colleagues and users easily has had a significantly positive impact on the learning experience for a large portion of our students. More and more researchers and academics are picking up this service as a convenient way to integrate open source stools which are essential to data science.
The main reason I personally changed over from Azure ML Studio is because it lacked any support for significant custom modelling with packages and services such as TensorFlow, scikit-learn, Microsoft Cognitive Toolkit and Spark ML. IBM Watson Studio provides these services and does so in a well integrated and easy to use fashion making it a preferable service over the other services that I have personally used.
IBM Watson Studio is very much suitable for data scientists when running a variety of analytical models using various languages such as R, Python and Scala. If you are planning to use data science driven languages in a cloud setting then IBM Watson Studio is a good option as it combines lots of relevant tools such as Notebooks, RStudio and Spark in a single environment. If you are looking to work in these environments as a group then Watson Studio also works well with the distribution and sharing of workspaces. This service however, isn't always the best solution as it can become costly if you are consistently running a large amount of intensive projects.

IBM Watson Studio Feature Ratings

Connect to Multiple Data Sources
8
Extend Existing Data Sources
8
Automatic Data Format Detection
7
MDM Integration
7
Visualization
9
Interactive Data Analysis
8
Interactive Data Cleaning and Enrichment
7
Data Transformations
8
Data Encryption
7
Built-in Processors
9
Multiple Model Development Languages and Tools
9
Automated Machine Learning
8
Single platform for multiple model development
8
Self-Service Model Delivery
7
Flexible Model Publishing Options
7
Security, Governance, and Cost Controls
8