What is document parsing?

Document parsing is the process of transforming documents into machine-readable formats such as text, HTML or Markdown, for down-stream tasks such as information search or summarization.

What approaches does Upstage provide?

At Upstage, we ensure that any types of documents, from scanned images to complex digital-born PDFs, can be processed with high quality and utilized in AI-driven workflows with the following approaches:

• Document parsing (Recommended): Detects and recognizes text along with document layout analysis, converting documents into structured, LLM-readable formats such as HTML or Markdown.
• Document OCR: Detects and recognizes text and their corresponding locations from any given document.

Note that document parsing leverages OCR under the hood, which means it is a more advanced technology and will provide additional capabilities such as understanding the document layout, complex tables, and charts.