Hi @Shashi123! You bring up a good point, the tutorial will not work on OCR out of the box. You need to run your PDFs through OCR tools that add a text layer, like Tesseract or Abbyy. However, reliable selectors will still be difficult to create.
A good alternative for OCRs is to use a cognitive data capture service to find data - it requires no explicit rule setup but rather uses AI to identify information, which is much better setup here. There’s a couple, one tutorial for UIPath to take a look at: https://rossum.ai/blog/2018/07/30/automating-data-extraction-from-invoices-using-rossum-api-and-uipath/