OCR Document Processing Improvements

Team, we currently have a BOT that was designed to read PDF invoices and validate 5 fields. Looking at the processing times using the default OCR technology within Studio we get the following ranges time ranges for the BOT.

30 secs to 65 seconds – after an error
25-45 seconds – after each invoice is submitted
18-25 seconds – for downloads

While I am not the developer of the BOT, he recommended looking at OCR improvements to help the BOT optimization.

Would you recommend a different OCR technology? Would you look at different parts of the code reading the invoice? Have you run into a similar issue when PDF processing times were running very slowly? Any help is appreciated. Thank you.

have you ever implemented with Intelligent OCR?