I created a new azure automl experiment and deployed it to an endpoint and can access the scoring URI via postman but how do I consume it in excel? Classic ml studio had the excel addin you can use but I don't see the same for URIs created and deployed from an automl experiment.
This Microsoft Developer video has a demo of exactly what I'm looking to do around the 32 min mark.
https://youtu.be/9FGuf55_Xtk?t=1915
@steves-4330 Thanks for the question, Have a look here:
https://github.com/retkowsky/AzureML_Excel
There is an Excel macro in the Excel file that call an Azure ML service deployed model.
There is a quick description of the process in the Word document available in this repo.
You can find here as well the Python notebook for creating & deploying the model. No autoML in it but not a big deal to adapt.
Please try the Consume web services portion Azure ML documentation? That could help you get started.
Thanks I was able to modify that macro to fit my models endpoint requirements. Hope we get another add in like ML classic studio had.
To make the macro work for my needs i had to edit the body that creates the JSON to send in the post and also the ret to return just the data i wanted in the return.
Hi Steve did it work for you? I love this "new" Azure ML studio but its integration with Excel is not good. Microsoft seems to nudge us into Power BI.
Also it would be good to see some examples of web services that work with an Azure Machine Learning model rest endpoint. I would like to see practical examples.
Thank you
6 people are following this question.
