Desktop PDF Data Extractor using Python

I want a Python-based engine that a non-technical user can download, install through a standard Windows installer, and then run entirely offline to process any PDF—no matter how large or complex. Once the user drags a file in, the program must extract the full set of Text, Tables, and Images, package the results into a clean, well-structured JSON file, and immediately sync that JSON to our own custom cloud endpoint (we already manage the endpoint; you simply need to post to it over HTTPS).

Everything about extraction happens locally: no external APIs for parsing or OCR. Feel free to rely on proven open-source libraries (pdfminer.six, PyPDF2, fitz/PyMuPDF, camelot or tabula-py for tables, Pillow or similar for image handling, and Tesseract if OCR becomes necessary), but bundle all dependencies so the end user never has to install Python. A simple GUI is welcome, yet a command-line toggle would also be helpful for batch automation.

Deliverables:
• The full extractor engine code with well-documented Python source
• Signed Windows installer (.msi or .exe) that drops a self-contained app folder
• JSON schema definition plus sample outputs covering text, table, and image cases
• Script or module that pushes the JSON to the provided cloud URL with basic retry logic
• Setup notes and build instructions so we can regenerate the installer in future

Acceptance will be based on accurate extraction across a test set of varied PDFs, valid JSON that conforms to the agreed schema, and successful sync to our cloud.