For the subset of PDFs that are invoices, there's the data extractor project called invoice2data, which implements many of the ideas presented in the article.

For the subset of PDFs that are invoices, there's the data extractor project called invoice2data, which implements many of the ideas presented in the article. https://github.com/invoice-x/invoice2data https://github.com/invoice-x/invoice2data