> ## Documentation Index
> Fetch the complete documentation index at: https://docs.raydocs.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Introduction to Raydocs

> AI-powered document extraction with full traceability

Raydocs is an AI-powered platform that automates the extraction of structured data from complex, unstructured documents. Every piece of extracted data comes with detailed reasoning and direct-to-source evidence, ensuring full **traceability** and **auditability**.

<CardGroup cols={2}>
  <Card title="Quick Start" icon="rocket" href="/guides/getting-started">
    Get up and running with your first extraction in minutes.
  </Card>

  <Card title="Integrate Raydocs with AI ✨" icon="sparkles" href="/guides/llm-quickstart">
    Give your AI assistant full documentation context.
  </Card>

  <Card title="Extraction Schema" icon="code" href="/guides/extraction-schema">
    Learn how to define extraction templates with the JSON schema.
  </Card>

  <Card title="API Reference" icon="book" href="/api-reference/introduction">
    Integrate Raydocs into your systems with our REST API.
  </Card>
</CardGroup>

## How It Works

Raydocs operates in two phases:

<Steps>
  <Step title="Document Processing">
    Documents are uploaded and processed through our RAG (Retrieval-Augmented Generation) pipeline. This includes parsing, intelligent chunking, and embedding generation to make your documents searchable.
  </Step>

  <Step title="Traceable Extraction">
    Using your extraction template, Raydocs retrieves relevant document chunks and extracts structured data. Every value includes reasoning and source references for complete transparency.
  </Step>
</Steps>

## Key Features

<AccordionGroup>
  <Accordion title="Full Traceability">
    Every extracted value includes:

    * **The extracted value** itself
    * **AI reasoning** explaining how the value was derived
    * **Source references** pointing to exact locations in the original document
  </Accordion>

  <Accordion title="Flexible Schema Definition">
    Define exactly what data you need using our JSON schema format:

    * Group related fields for efficient extraction
    * Use field references and dependencies
    * Support for arrays, nested objects, and reusable definitions
  </Accordion>

  <Accordion title="AI-Powered Document Processing">
    Documents are processed using Vision Language Models (VLM) for optimal extraction:

    * Handles all document types including PDFs, presentations, and images
    * Superior understanding of visual layouts and complex formatting
    * Optional contextual embedding for enhanced search quality
  </Accordion>

  <Accordion title="Export & Integration">
    Get your data where you need it:

    * Export to Excel with full audit trails
    * REST API for system integration
    * Batch processing for high-volume workflows
  </Accordion>
</AccordionGroup>

## Use Cases

Raydocs excels at extracting structured data from:

* **Financial Reports**: Revenue figures, KPIs, investment data
* **Legal Documents**: Contract terms, party information, key dates
* **Research Papers**: Findings, methodologies, citations
* **Corporate Filings**: Company details, officer information, regulatory data

## Getting Started

Ready to start extracting data? Head to the [Quick Start Guide](/guides/getting-started) to create your first extraction template and process your first document.