Skip to main content
SAPPatent grant

SAP was granted a patent for stronger document information extraction.

Source: Autobound Signal API · 149 signals tracked for SAP · updated

Get SAP's live signals in Claude or ChatGPT, or by API

AskWhat changed at SAP this week? Use Autobound signals for sap.com.

What happened

In November 2026 SAP patented a training approach for extracting business data from invoices, forms, and other scanned or digital documents.

Source

patents.google.comNov 17, 2026

US patent US12802481

Virtual adversarial training for document information extraction models

Abstract

The present disclosure relates to computer-implemented methods, software, and systems for extracting information from documents based on training techniques to generate a document foundation model that is used to initialize a document information extraction model that is fine-tuned to business document specifics. A document information extraction model is initialized based on weights provided from a first pretrained model. Fine-tuning of the document information extraction model is performed based on labeled business documents as second training data. The labeled business documents are labeled and evaluated according to a virtual adversarial training (VAT). Based on the performed fine-tuning, a classifier for classification of information extraction is generated.

patents.google.com/patent/US12802481Read the full source

Extracted by Autobound

From the Signal API record
Signal
Patent grant

What this signalsA new patent often shows where a company is putting its R&D budget.

Grant date
Nov 17, 2026
Inventors
Christoph Batke, Sohyeong Kim
Tech area
computer vision
Patent number
US12802481

The full record

From the Signal API record

People

  • Christoph BatkeInventor
  • Sohyeong KimInventor

Details

Primary CPC class
G06V 30/413 (Image or video recognition or understanding)
USPTO assignee
SAP SE
What the invention does
A method that improves document-reading models by starting with broad training and fine-tuning them with deliberately difficult variations of business documents.

Topics and mentions

Tags

  • Intellectual Property
  • Research & Development
  • Artificial Intelligence
  • Machine Learning
  • Enterprise Software

Technologies named

  • document AI
  • computer vision
  • machine learning
  • information extraction

Extraction

Detected
Nov 18, 2026
signal_type
patents-company
signal_subtype
patentGrant

Use this data

Get every patent signal for SAP and the companies you sell to, in the tools you already use.

  1. Ask Claude about it

    Connect Autobound to Claude, Claude Code or Cursor with MCP. Then ask: “What changed at SAP this week?”

  2. Send it to your own tools

    The Signal API returns patent signals for any list of companies as JSON, for your CRM, warehouse or app.

  3. Try it free

    Sign up and spend your free credits on the companies you sell to.

    Start Free1,000 free credits

The API returns more than this page shows

This page shows a preview. The full patents-company record in the Signal API and MCP can also have these 8 fields. Some fields are empty for some signals.

Company

  • linkedin_urlValue in the API
  • industriesValue in the API
  • employee_count_lowValue in the API
  • employee_count_highValue in the API
  • revenueValue in the API
  • descriptionValue in the API

Signal

  • signal_nameValue in the API
  • associationValue in the API
Show the full JSONThe record on this page and the API request

GET /v1/signals/10ac8d2a-9993-565d-9784-116094243da1 returns this record as JSON. POST /v1/companies/enrich returns every signal for sap.com.

{
  "signal_id": "10ac8d2a-9993-565d-9784-116094243da1",
  "signal_type": "patents-company",
  "signal_subtype": "patentGrant",
  "detected_at": "2026-11-18T15:39:56+00:00",
  "company": {
    "name": "SAP",
    "domain": "sap.com"
  },
  "data": {
    "cpc": [
      {
        "code": "G06V 30/413",
        "label": "Image or video recognition or understanding"
      },
      {
        "code": "G06V 30/19073",
        "label": "Image or video recognition or understanding"
      },
      {
        "code": "G06V 30/1916",
        "label": "Image or video recognition or understanding"
      },
      {
        "code": "G06V 30/19173",
        "label": "Image or video recognition or understanding"
      }
    ],
    "tags": [
      "Intellectual Property",
      "Research & Development",
      "Artificial Intelligence",
      "Machine Learning",
      "Enterprise Software"
    ],
    "detail": "In November 2026 SAP patented a training approach for extracting business data from invoices, forms, and other scanned or digital documents.",
    "summary": "SAP was granted a patent for stronger document information extraction.",
    "event_at": "2026-11-17",
    "inventors": [
      {
        "name": "Christoph Batke",
        "status": "resolved",
        "is_primary": true,
        "linkedin_url": "https://www.linkedin.com/in/christoph-batke-450771194"
      },
      {
        "name": "Sohyeong Kim",
        "status": "resolved",
        "is_primary": false,
        "linkedin_url": "https://www.linkedin.com/in/sohyeong-kim-bb472775"
      }
    ],
    "tech_area": "computer vision",
    "event_kind": "grant_date",
    "grant_date": "2026-11-17",
    "source_url": "https://patents.google.com/patent/US12802481",
    "cpc_primary": {
      "code": "G06V 30/413",
      "label": "Image or video recognition or understanding"
    },
    "patent_title": "Virtual adversarial training for document information extraction models",
    "patent_number": "US12802481",
    "uspto_assignee": "SAP SE",
    "patent_abstract": "The present disclosure relates to computer-implemented methods, software, and systems for extracting information from documents based on training techniques to generate a document foundation model that is used to initialize a document information extraction model that is fine-tuned to business document specifics. A document information extraction model is initialized based on weights provided from a first pretrained model. Fine-tuning of the document information extraction model is performed based on labeled business documents as second training data. The labeled business documents are labeled and evaluated according to a virtual adversarial training (VAT). Based on the performed fine-tuning, a classifier for classification of information extraction is generated.",
    "precision_class": "structural",
    "invention_explanation": "A method that improves document-reading models by starting with broad training and fine-tuning them with deliberately difficult variations of business documents.",
    "technologies_mentioned": [
      "document AI",
      "computer vision",
      "machine learning",
      "information extraction"
    ]
  }
}

Long text fields are shortened on this page.

Looking up one signal by its id is free. Enrich costs 2 credits per signal returned; a call with no results is free.