I’m designing a web-based document OCR system and would like advice on the appropriate frontend, backend, database, and deployment setup.

The system will be hosted and will support two user roles: a general user who uploads documents and reviews OCR results, and an admin who manages users and documents.

There are five document types. Two document types have varying layouts, but I only need to OCR the person’s name and the document type so it can be matched to the uploader. One document type follows a two-column key–value format such as First Name: John. For this type, I need to OCR both the field label and its value, then allow the user to manually correct the OCR result if it is inaccurate. The remaining document types follow similar structured patterns.

For the frontend, I am most familiar with React.js and Next.js. I prefer using React.js with shadcn/ui for building the UI and handling user interactions such as file uploads and OCR result editing.

For the backend, I am considering FastAPI to handle authentication, file uploads, OCR processing, and APIs. For my OCR, I am thinking of using PaddleOCR but I am also open to other recommendations. And also searching for other OCR tools for my usecase.

My main questions are:

* Is React.js with shadcn/ui a good choice for this type of application, or would Next.js provide meaningful advantages?
* Is FastAPI suitable for an OCR-heavy workflow that includes file uploads and asynchronous processing?
* Are there known deployment or scaling issues when using Next.js (or React) together with FastAPI?
* What type of database would be recommended for storing users, document metadata, OCR results, and corrected values?

I’m trying to avoid architectural decisions that could cause issues later during deployment or scaling, so insights from real-world experience would be very helpful.

Thanks in advance.