Dear Everyone I am facing a challenge from a client, the client has asked for Realtime OCR processing, there are 6 document fields and the requirement is to fetch dates, number etc from those 6 documents and populate the netsuite date and freefromtext field. The documents can be PDF, Image and Word, etc. there can be 14 page long pdf and there can be a single image. I opted document AI for OCR and I have 1 working way for Realtime but for that user will have to wait for 5 minutes and continuously press refresh button another way is to achieve it after record save, in this way i will make gcp cloud functiion as webhook but the client want it in realtime when the user creates record/edit record and want the processing time of few seconds Do you guys have any better idea or past experience, Please share it will be very helpful Thanks

Isn't docAI itself pretty quick? What is causing the 5 mins delay?

docAI took 2.5 minutes for processing a pdf having number of 14 pages although i added parameter, pages: [1,2,3]

Interesting, a colleague did a little live demo of doc AI a few weeks and it maybe took a few-5 seconds it was pretty quick. Not sure If I can offer much help. Is the workflow something that really needs to happen live? And what is being extracted from the 5-6 docs/what is the action after the data is extracted

Can't help with speeding up the API, but you can improve UX by using a client script to poll for OCR status and show a waiting indicator. Once it's done, notify the user automatically instead of requiring them to constantly refresh manually