If you download the Microsoft Intelligent Services v6.2 from the document, there's a PDF which shows how to setup to read images using OCR. Follow the steps exactly and it should work. The additional info that is not mentioned in the user guide is that the Microsoft Cognitive Services VBO has hard-coded the API URL to be from North Europe. So when you create the resource in Azure, pls use the same location. It should then work. And pls note that the OCR will only read from images, not PDF. So you'll also need to convert your PDF to image first.
