azure-ai-document-intelligence-ts

What problem does it solve?

Developers and automation engineers can reliably extract text, tables, and structured fields from documents without manual transcription or fragile OCR pipelines by leveraging Azure Document Intelligence prebuilt and custom models.

Core Features & Use Cases

• Prebuilt model extraction: Use prebuilt-read, prebuilt-layout, prebuilt-invoice, prebuilt-receipt, and other models to extract text, tables, and domain fields.
• Flexible inputs: Analyze documents from URLs or local files encoded as base64 and process single documents or batch jobs to a result container.
• Model management & classification: Build custom document models and classifiers for specialized document types and list or manage models programmatically.
• Operational patterns: Uses long-running pollers for asynchronous operations, helpers for pagination and artifact retrieval, and explicit error checking with isUnexpected.
• Use Case: Automate invoice and receipt processing to extract vendor names, line items, totals, and due dates into structured outputs for downstream accounting workflows.

Quick Start

Analyze the invoice at https://example.com/invoice.pdf using the prebuilt-invoice model and return extracted fields, tables, and confidence scores.