Analytics Legends The knowledge platform for SAP Analytics
Academy module

SAP Document AI — Document Automation End to End

SAP Document AI — Document Automation End to End — Analytics Legends section illustration for the SAP Analytics knowledge base (concepts, studies, Academy)

As of 2026-09-23

Edited by , sole editor and publication director — 27 years in the SAP ecosystem. How these pages are produced, and where automation is used.

SAP Document AI (formerly Document Information Extraction) turns invoices, orders, payment advices and custom documents into structured data. This module teaches the four enterprise plans (base, premium, embedded — charged in AI units — and free), schema and setup-type design (ML model, generative AI or template per field), the REST and OData APIs, SAP Build Process Automation integration including the destination-based premium pattern, confidence-based human review, and the July–September 2026 changes.

What you will learn

  • Map a customer's document types to standard or custom schemas and choose, per field, the extraction method (SAP ML model, generative AI or template) from the setup-type table
  • Select the right Document AI service plan (base, premium, embedded, free) and explain its metric — blocks of 100 documents, compute hours, requests, users or AI units
  • Run a first extraction through the basic UI and the /document/jobs REST API and record a per-field confidence baseline
  • Integrate extraction into SAP Build Process Automation with built-in activities (base) or a destination-based package (premium)
  • Design confidence-based human validation on mandatory fields, with an audit trail and Instant Learning feedback

Module overview

SAP Document AI is the SAP Business Technology Platform (SAP BTP) solution that turns business documents — invoices, purchase orders, payment advices, business cards and any custom document you define — into structured data that a process, an API or a human reviewer can act on. It is the product that SAP previously called Document Information Extraction (DOX). The rename matters in practice: SAP Help for Document AI uses the new name, but the SAP Build Process Automation (SBPA) help pages, the SDK package and many API paths still say "Document Information Extraction (former name)" and /document-information-extraction/v1. A senior consultant must be fluent in both vocabularies, because the customer's licence contract, the BTP cockpit and the automation designer will not all use the same word.

Prerequisites

  • Working knowledge of SAP BTP subaccounts, entitlements, service instances and service keys
  • Review core concepts first: C332, C026, C087
  • Basic SAP Build Process Automation: processes, automations, decisions and My Inbox

Outcomes

  • Explain to a buyer which Document AI edition delivers generative-AI extraction and how each edition is metered, without quoting unverified prices.
  • Produce a field-level schema design with setup type, extraction method and mandatory-field flag for one standard and one custom document.
  • Build an SBPA flow that extracts, routes low-confidence documents to My Inbox and posts confirmed values.
  • Diagnose a failing extraction by comparing the called endpoint with the service-key URL before changing entitlements.

Full module available to members. The full module adds: the decision framework · the end-to-end scenario walkthrough · the KPI scorecard · the anti-patterns · the code blocks · the knowledge check · the diagrams.

Open in the app →

Next in this track