Multi-Source machine learning done well.
December 08, 2020

Multi-Source machine learning done well.

Ben Williams | TrustRadius Reviewer
Score 9 out of 10
Vetted Review
Verified User

Overall Satisfaction with SAP Data Intelligence

SAP Data Intelligence has been used specifically within my department as a means of predicting certain costs of a claim. This has been an on-going project taking data from an OCR engine which reads incoming documents passing details to a variety of databases and SAP in order to predict those claims as possibly fraudulent claims or high value claims and pass this information to different departments via report for them to "keep an eye on."
  • Easy to pick up.
  • Tools for data cleansing.
  • Able to use existing learning models from the business.
  • Operates on cloud.
  • Not made for working with others.
  • Debugging is needlessly complicated and obfuscates issues.
  • Documentation could use more examples to highlight points.
  • Streamlined the claims process.
  • Savings made on claims as those high value ones have been highlighted faster.
  • Saved on analyst time as the tool has eased some of the workload by tying together different elements.
We used SAP for a variety of data tasks already and were approached with this software to assist with our decision making. I've not been involved in the tendering process but I know we had a few different options and decided this worked best with our current suite of software tools.

Do you think SAP Data Intelligence delivers good value for the price?

Yes

Are you happy with SAP Data Intelligence's feature set?

Yes

Did SAP Data Intelligence live up to sales and marketing promises?

I wasn't involved with the selection/purchase process

Did implementation of SAP Data Intelligence go as expected?

Yes

Would you buy SAP Data Intelligence again?

Yes

SAP DI has proved useful to us with its ability to pull from a variety of data sources and is well suited to this. It is also more than capable of cleansing out data without a great deal of issues. It is well suited to running a variety of different ML models including the use of python and R. It is not so well suited to large masses of data and can do with more support around complex ML models.