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

# Introduction

> The foundation layer for drawing intelligence. Parse construction drawings into a structured knowledge graph and query it from your app or AI agent.

## What is Bedrock?

Bedrock is the foundation layer for construction drawing intelligence. Upload PDF drawing sets and Bedrock's vision pipeline segments every page into sheets and blocks, classifies content, extracts text, and indexes everything for search. Use the Vision API to detect symbols, generate revision overlays, and analyze changes.

* **Parse any drawing** — Architectural, structural, MEP, civil. Automatic preprocessing splits PDFs into sheets and blocks.
* **Detect symbols** — Legend-driven symbol detection finds and counts fixtures, equipment, and annotations across plan views.
* **Compare revisions** — Generate visual overlays between drawing revisions and analyze what changed.
* **Search and ask** — Hybrid semantic + visual search across your project, with natural language Q\&A.
* **Connect AI agents** — Give Claude, ChatGPT, or Microsoft Copilot direct access via the MCP server.

<CardGroup cols={2}>
  <Card title="Quick Start" icon="rocket" href="/guides/quickstart">
    Upload a drawing and query the Drawing Index in 5 minutes.
  </Card>

  <Card title="Glossary" icon="spell-check" href="/guides/glossary">
    New to construction? AEC terminology explained.
  </Card>
</CardGroup>

## The Drawing Index

The Drawing Index is a structured knowledge graph of everything in your construction drawing set. When you upload a PDF, Bedrock parses it into a hierarchy of entities with typed relationships between them:

```mermaid theme={null}
graph TD
  D["Drawing (PDF)"]:::drawing --> S["Sheet (A-101)"]:::sheet
  S --> B1["Plan view"]:::block
  S --> B2["Door Schedule"]:::block
  S --> B3["Legend"]:::block

  B1 --> F1["room '201'"]:::feature
  B1 --> F2["door '104'"]:::feature
  B1 --> F3["duplex_receptacle"]:::feature

  F2 -. "scheduled_in" .-> B2
  F3 -. "defined_by" .-> B3

  classDef drawing fill:#009689,color:#fff,stroke:#007a6f
  classDef sheet fill:#00BBA7,color:#fff,stroke:#009689
  classDef block fill:#e0f2f1,color:#333,stroke:#009689
  classDef feature fill:#f5f5f5,color:#555,stroke:#bbb
```

**Drawings** contain **Sheets** (pages). Each Sheet contains **Blocks** (plan views, schedules, legends). Each Block contains **Features** (rooms, doors, symbols). **Relations** connect entities across the graph — a door to its schedule row, a symbol to its legend definition, a callout to a detail on another sheet. A shared **Grid** coordinate system aligns everything across disciplines.

<Card title="The Drawing Index" icon="book" href="/guides/the-drawing-index">
  Deep dive into the knowledge graph and how queries work.
</Card>

## APIs

Bedrock exposes two REST APIs. The **CMS API** is your CRUD interface for managing projects and reading every entity in the Drawing Index. The **Vision API** triggers processing: detecting symbols, generating overlays, and analyzing changes.

<CardGroup cols={2}>
  <Card title="CMS API" icon="database" href="/cms-api/introduction">
    Projects, drawings, sheets, blocks, features, relations, overlays. Standard REST with filtering, pagination, and expansion.
  </Card>

  <Card title="Vision API" icon="eye" href="/vision-api/introduction">
    Detect symbols, generate revision overlays, analyze changes. Async jobs with polling and streaming. Synchronous search and Q\&A.
  </Card>
</CardGroup>

|             | CMS API                   | Vision API       |
| ----------- | ------------------------- | ---------------- |
| **Purpose** | Read and manage data      | Trigger analysis |
| **Pattern** | Synchronous REST          | Async jobs       |
| **Credits** | Free                      | Per-operation    |
| **Example** | `GET /features?type=door` | `POST /parse`    |

## MCP Server

The MCP (Model Context Protocol) server gives AI agents direct access to the Drawing Index. Connect Claude, ChatGPT, Microsoft Copilot, or any MCP-compatible client and ask natural language questions about your drawings.

```json theme={null}
{
  "mcpServers": {
    "bedrock": {
      "url": "https://mcp.bedrock.cv/mcp",
      "headers": { "Authorization": "Bearer sk_your_api_key" }
    }
  }
}
```

Eight composable tools handle any construction drawing query:

| Tool        | Purpose                                                                 |
| ----------- | ----------------------------------------------------------------------- |
| `Query`     | Get/list/filter entities (projects, drawings, sheets, blocks, features) |
| `Search`    | Hybrid semantic + keyword search across project content                 |
| `Ask`       | Natural language Q\&A with source citations                             |
| `Compare`   | Trigger revision comparison                                             |
| `Parse`     | Trigger symbol detection                                                |
| `PollJob`   | Poll async job status and results                                       |
| `ViewImage` | Retrieve block or feature images                                        |
| `SQL`       | Read-only SQL queries for analytics                                     |

<CardGroup cols={2}>
  <Card title="MCP Setup" icon="plug" href="/mcp/setup">
    Connect the Drawing Index to your AI agent in 2 minutes.
  </Card>

  <Card title="Workflow Examples" icon="route" href="/mcp/workflows/symbol-counting">
    See how agents chain tools to count symbols, compare revisions, and more.
  </Card>
</CardGroup>