Seeking Solutions for Programmatic Bates Numbering and eIDAS-Compliant Signing in PDFs

I'm working on a .NET application (leveraging the Azure ecosystem) that automates legal document processing. I've hit a couple of specific roadblocks with PDF manipulation that I'm hoping someone here has experience with.

My core requirements are:

Dynamic Bates Numbering: I need to apply sequential Bates numbers (e.g., LXN_0000001) to large sets of documents—often thousands of pages. The challenge is that this needs to be done programmatically via an API/SDK, and the numbering must be searchable, selectable text (not just an image stamp) and ideally embedded as metadata. Many libraries I've tested only offer basic watermarking, which isn't sufficient for e-discovery standards.

eIDAS-Compliant Digital Signatures: The final PDFs often need to be signed with digital signatures that are compliant with EU eIDAS regulations (Advanced Electronic Signatures). This requires the signature to be cryptographically tied to the document with a qualified certificate and a secure audit trail, ensuring non-repudiation and long-term validity (LTV).

I've explored some open-source options, but found the implementation for these specific, high-trust requirements to be complex, time-consuming, and risky from a compliance perspective.

My question is: Has anyone successfully implemented a similar workflow for high-stakes environments like legal or finance? Can you recommend a robust API or SDK that can handle these tasks reliably without sacrificing security or compliance? Any guidance or shared experiences would be immensely helpful.

Thanks in advance for your insights!