Oracle Cloud Infrastructure Documentation

Pretrained Document AI Models

Vision provides pretrained document AI models that allow you to organize and extract text and structure from business documents.

Pretrained models let you use AI with no data science experience. Simply provide an image-based document to the Vision service and get back information about your document without having to create your own model.
Important

The AnalyzeDocument and DocumentJob capabilities in Vision are moving to a new service, Document Understanding. The following features are impacted:
  • Table detection
  • Document classification
  • Receipt key-value extraction
  • Document OCR
These features are available in Vision until January 1, 2024. After then, they are available only in Document Understanding.

Use Cases

Pretrained document AI models let you automate back-office operations, and process receipts more accurately.

Intelligent search
Enrich image-based files with metadata, including document type and key fields, for easier retrieval.
Expense reporting
Extract the required information from receipts to automate business workflows. For example, employee expense reporting, spending compliance, and reimbursement.
Downstream Natural Language Processing (NLP)
Extract text from PDF files and organize it as the input for NLP, either in tables or in words and lines.
Loyalty points capture
Automate loyalty points calculations from receipts, based on the number of items or the total amount paid.

Supported Formats

Vision supports several document formats.

Documents can be uploaded either from a local file or Oracle Cloud Infrastructure Object Storage. They can be in the following formats:
  • JPEG
  • PDF
  • PNG
  • TIFF