Introduction > Introduction

Introduction

Welcome to AI Data Foundry.

AI Data Foundry is a powerful AI-based tool that extracts text, layout, tables, and key-value pairs from various unstructured document formats including HWP, Docx, and PDF.

The extracted data can be immediately utilized in various workflows such as RAG (Retrieval-Augmented Generation) data preprocessing, RPA (Robotic Process Automation), and AI dataset construction.

Key Features

Maximize data utilization with diverse document support and accurate document structure analysis

Support for Various Business Document Formats

Supports a wide range of document formats including Hangul / MS-Office / PDF / Images, enabling compatibility with most document types held by enterprises

Perfect Analysis of Hidden Document Structure

  • Recognition of detailed document structure information including headings, paragraphs, headers, footers, page numbers, captions, lists, etc.
  • Recognition of visual information such as tables and images

Highly Practical Output Formats

  • Markdown output support for Large Language Model (LLM) construction
  • XML output support for enterprise database construction

Use Cases

LLM Model Training

Understanding various types of documents and
building LLM models based on owned documents

Conversational AI (RAG)

Improving natural language processing capabilities through understanding documents including tables and images
and developing conversational AI Q&A systems

Business Automation (RPA)

Extracting only necessary information to
build business automation systems

Digital Archive

Structuring unstructured documents and
building large-scale document digital archives

Next Steps

See the basic usage flow in Getting Started.