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

# Introduction

> Extraction and processing engines for structured data transformation

## What are Engines?

Engines are specialized processing tools designed for specific extraction and transformation tasks. Unlike [Agentic Workflows](/agents/agentic-workflows/introduction) that reason and investigate, engines follow deterministic processes to extract, parse, and structure data from documents, images, and web content.

### When to Use Engines

Choose an engine when you need:

* **Fast, predictable extraction** from PDFs, images, or web pages
* **Specific transformations** like document segmentation or URL finding
* **Structured output** from unstructured sources
* **Repeatable processing** with consistent results

For complex investigations requiring reasoning, research, and decision-making, see [Agentic Workflows](/agents/agentic-workflows/introduction).

## Available Engines

<CardGroup>
  <Card title="Document Insights Engine" icon="file-pdf" href="/agents/engines/pdf-extraction">
    Extracts insights and information from PDF documents into structured JSON.
  </Card>

  <Card title="Document Segmentation Engine" icon="scissors" href="/agents/engines/document-segmentation">
    Selects and extracts specific pages from PDF documents.
  </Card>

  <Card title="Google Maps Insights Engine" icon="map-location-dot" href="/agents/engines/google-maps">
    Extracts location information and nearby places from Google Maps.
  </Card>

  <Card title="Interactive Web Insights Engine" icon="browser" href="/agents/engines/interactive-web">
    Performs interactive web browsing and extracts information from dynamic pages.
  </Card>

  <Card title="Marketplace Storefront Analysis Engine" icon="store" href="/agents/engines/marketplace-storefront">
    Analyzes marketplace storefronts for compliance and quality.
  </Card>

  <Card title="Merchant Risk Analysis Engine" icon="building" href="/agents/engines/merchant-risk">
    Analyzes merchant risk profiles for underwriting and compliance.
  </Card>

  <Card title="Multimodal Insights Engine" icon="images" href="/agents/engines/multimodal-extraction">
    Extracts structured information from text and images using AI vision models.
  </Card>

  <Card title="Research Agent Engine" icon="magnifying-glass" href="/agents/engines/research">
    AI-powered web research to gather and synthesize information.
  </Card>

  <Card title="Social Media Crawler Engine" icon="share-nodes" href="/agents/engines/social-media-crawler">
    Scrapes profile information from social media platforms.
  </Card>

  <Card title="URL Finder Engine" icon="link" href="/agents/engines/url-finder">
    Discovers and extracts URLs from web pages and content.
  </Card>

  <Card title="Web Insights Engine" icon="globe" href="/agents/engines/url-extraction">
    Extracts content and information from web pages.
  </Card>
</CardGroup>

## Engine Configuration Overview

All engines come with their own configuration settings. The settings defines parameters like which AI model or prompts the Agent will use when running jobs. If your configurations need to be dynamic (changes depending on the job), then you can add as many Agent inputs as you need and use them in the Engine configuration. Check out [Template Strings](/agents/input-definition#template-strings) to see how to do that.