Dec 20, 2016
Azure ML is a bit slow. Training on 50,000 data set toom 15 minutes. The data set was housed in the Azure Blob. We also rean into SWAGGER error, which we needed to reach out to Microsoft vender to help trouble shoot.
Dec 6, 2016
The overall experience was good. Having a strategic relationship with MS enables us to have access to resources that typically would not be available to all other organizations. With that the ability to ask the right questions and have Design Sessions enabled us to understand the capability and have a better understanding of how the capability was going to be leveraged.
Dec 1, 2016
All worked very well. Minor and few things did not work very well but they were very managable.
Nov 29, 2016
For Analysis Services the research support is relatively limited
Nov 11, 2016
Great web-based UI for ETL, modeling, and deployment to web service. Some challenges remain around versions of open source packages supported, but overall the platform is best in class.
Oct 24, 2016
We moved the 90% of all our internal machine learning systems to the Azure ML Studio in less than 3 weeks. Now all the process is in strreaming in Azure and we have the possibility to lunch multiply instance of test / learning with less than 60% of the previous time and resources. Using Python and Jupiter notebook integration!
Oct 20, 2016
Ease to implement the dificult piece was changing clien culture totake decisions based om information
Oct 19, 2016
Oct 18, 2016
Still reviewing and testing product due to so many internal legacy systems we would need to ensure that any product utilized maintains no risk because of PII data.
Oct 17, 2016
Azure ML very easy to pickup. We had a number of different profiles of people learn Azure ML (Business Analysts, Data Analysts), even summer interns. They were able to quickly understand the tool and use it to solve business problems.