Account Data PDF - OCR

Hello,
trying to extract data with from the Annual Account Files (PDF) for companies that do not deliver an ixbrl-account file.

I am currently having a try with https://github.com/ONSBigData/parsing_company_accounts

There are coming some results, but also a lot of reproducable fails due to runtime errors within the workflow for specific pdf-files. But seems to work better for me than this repo:

Any other repository suggestions I may try?