SAP Document AI – Integration from Multiple External Web Systems (PDF Upload) & Response Data / Rete

[#1]

Hi all,

We are currently designing an integration architecture involving SAP Document AI and would appreciate some clarification from the community or SAP experts.

We operate multiple external web systems, and users upload PDF documents through each of these systems. We are evaluating whether these PDFs can be sent to SAP Document AI for processing (OCR / data extraction), as part of a broader automation/AI transformation initiative we're currently designing.

### Questions

**1. Feasibility**

Is it possible for PDF documents uploaded from multiple external web systems to be sent to SAP Document AI via API for processing (OCR/data extraction)?

**2. Response data and reusability**

When calling the API, what data is returned in the response (e.g., extracted fields, confidence scores, whether the original document is included, etc.)?

- Is this response meant to be consumed immediately by the calling system only (synchronous, one-time use)?
- Or is the extraction result persisted on the SAP Document AI side, so that it can be retrieved later by a different system or at a different point in time (i.e., queried again after the fact)?

**3. Document retention**

Is the original PDF file itself stored on the SAP side after being submitted? If so, how long is it retained (e.g., indefinitely, or deleted after a certain period)?

---

Any insight — whether from documentation, hands-on experience, or official SAP guidance — would be greatly appreciated.

Thanks in advance!