RAAS API Documentation

Integrate document intelligence into your agents, pipelines, and applications

Authentication

All API requests require a Bearer token. Generate an API key from your dashboard. Keys have the format raas_...

Base URL: https://api.raas.tools

curl -H "Authorization: Bearer raas_your_api_key" \
  https://api.raas.tools/api/documents

Documents API

POST/api/documents

Upload a document for processing. Send as multipart/form-data with the file and a project ID.

Request body (multipart/form-data)

`file`	File	The document file (PDF, DOCX, XLSX, PPTX, TXT)
`projectId`	string	UUID of the target project

curl -X POST https://api.raas.tools/api/documents \
  -H "Authorization: Bearer raas_your_key" \
  -F "file=@contract.pdf" \
  -F "projectId=your-project-id"

Response

{
  "id": "doc_abc123",
  "name": "contract.pdf",
  "status": "queued",
  "projectId": "your-project-id",
  "createdAt": "2026-03-15T10:30:00.000Z"
}

GET/api/documents

List all documents in a project. Returns metadata and processing status for each document.

Query parameters

projectId string UUID of the project (required)

curl "https://api.raas.tools/api/documents?projectId=your-project-id" \
  -H "Authorization: Bearer raas_your_key"

Response

{
  "data": [
    {
      "id": "doc_abc123",
      "name": "contract.pdf",
      "status": "completed",
      "entityCount": 24,
      "chunkCount": 87,
      "createdAt": "2026-03-15T10:30:00.000Z"
    }
  ]
}

GET/api/documents/:id/progress

Get real-time processing status for a document. Poll this endpoint after upload to track ingestion progress.

curl https://api.raas.tools/api/documents/doc_abc123/progress \
  -H "Authorization: Bearer raas_your_key"

Response

{
  "status": "processing",
  "processingStep": "embedding",
  "processingProgress": 0.72,
  "entityCount": 24,
  "chunkCount": 65
}

Search API

POST/api/search/hybrid

Combined semantic + keyword search with optional Cohere reranking. This is the recommended search endpoint for most use cases.

Request body

{
  "query": "What are the payment terms?",
  "projectId": "your-project-id",
  "limit": 5,
  "rerank": true
}

Response

{
  "data": [
    {
      "chunk": {
        "text": "Payment is due within 30 days of invoice date...",
        "metadata": { "page": 4, "section": "Terms" }
      },
      "score": 0.94,
      "document": {
        "id": "doc_abc123",
        "name": "contract.pdf"
      }
    }
  ]
}

POST/api/search/semantic

Pure vector similarity search using pgvector embeddings. Best for open-ended or conceptual queries.

{
  "query": "risks related to supply chain disruption",
  "projectId": "your-project-id",
  "limit": 10
}

GET/api/search/entities

Query extracted entities (people, dates, monetary amounts, etc.) with structured filters. Uses SQL under the hood for exact matching.

Query parameters

`projectId`	string	Project UUID (required)
`type`	string	Entity type filter (e.g. "amount", "date", "person")
`minConfidence`	number	Minimum confidence score (0-1, default 0.7)

curl "https://api.raas.tools/api/search/entities?projectId=ID&type=amount&minConfidence=0.8" \
  -H "Authorization: Bearer raas_your_key"

Response

{
  "data": [
    {
      "type": "amount",
      "value": "$3,200,000",
      "confidence": 0.95,
      "context": "Q4 Enterprise SaaS revenue totalled $3,200,000",
      "document": { "id": "doc_abc123", "name": "q4-report.pdf" }
    }
  ]
}

Chat API

POST/api/chat/conversations/:id/messages

Send a message to a document conversation. Returns an AI-generated response with source citations.

Request body

{
  "content": "What were the top revenue streams in Q4?"
}

Response

{
  "message": {
    "role": "assistant",
    "content": "The top three revenue streams in Q4 were Enterprise SaaS ($3.2M), Professional Services ($1.1M), and API Usage ($440K)."
  },
  "citations": [
    {
      "text": "Q4 Enterprise SaaS revenue totalled $3,200,000",
      "document": "q4-report.pdf",
      "page": 7
    }
  ]
}

MCP Integration

RAAS ships as a Model Context Protocol server. Connect it to Claude Desktop, Cursor, or any MCP-compatible client to give your AI tools direct access to your documents.

claude_desktop_config.json

{
  "mcpServers": {
    "raas": {
      "command": "npx",
      "args": ["-y", "@raas/mcp"],
      "env": {
        "RAAS_API_KEY": "raas_your_key",
        "RAAS_PROJECT_ID": "your-project-id"
      }
    }
  }
}

Available MCP Tools

`search_documents`	Semantic search over your document corpus
`hybrid_search`	Combined keyword + semantic search with reranking
`query_entities`	Query extracted entities (people, dates, amounts, etc.)
`compare_documents`	Analyze differences between two documents
`extract_to_format`	Extract data into a structured schema
`generate_report`	Generate a structured report from documents
`check_compliance`	Validate content against rules
`draft_from_templates`	Generate drafts using document templates

Rate Limits

Tier	API Calls	Queries/mo	Documents
Free	60/min	500	100
Pro	100/min	2,500	500
Enterprise	Custom	50,000	10,000

When rate limited, the API returns 429 with a Retry-After header.

Quick Start

A complete working example — upload a document then search it.

# 1. Upload a document
curl -X POST https://api.raas.tools/api/documents \
  -H "Authorization: Bearer raas_your_key" \
  -F "file=@contract.pdf" \
  -F "projectId=your-project-id"

# 2. Search across your documents
curl -X POST https://api.raas.tools/api/search/hybrid \
  -H "Authorization: Bearer raas_your_key" \
  -H "Content-Type: application/json" \
  -d '{
    "query": "What are the payment terms?",
    "projectId": "your-project-id",
    "limit": 5,
    "rerank": true
  }'

RAAS API Documentation

Integrate document intelligence into your agents, pipelines, and applications

Get Started View Pricing

Authentication

All API requests require a Bearer token. Generate an API key from your dashboard. Keys have the format raas_...

Base URL: https://api.raas.tools

curl -H "Authorization: Bearer raas_your_api_key" \
  https://api.raas.tools/api/documents

Documents API

POST/api/documents

Upload a document for processing. Send as multipart/form-data with the file and a project ID.

Request body (multipart/form-data)

`file`	File	The document file (PDF, DOCX, XLSX, PPTX, TXT)
`projectId`	string	UUID of the target project

curl -X POST https://api.raas.tools/api/documents \
  -H "Authorization: Bearer raas_your_key" \
  -F "file=@contract.pdf" \
  -F "projectId=your-project-id"

Response

{
  "id": "doc_abc123",
  "name": "contract.pdf",
  "status": "queued",
  "projectId": "your-project-id",
  "createdAt": "2026-03-15T10:30:00.000Z"
}

GET/api/documents

List all documents in a project. Returns metadata and processing status for each document.

Query parameters

projectId string UUID of the project (required)

curl "https://api.raas.tools/api/documents?projectId=your-project-id" \
  -H "Authorization: Bearer raas_your_key"

Response

{
  "data": [
    {
      "id": "doc_abc123",
      "name": "contract.pdf",
      "status": "completed",
      "entityCount": 24,
      "chunkCount": 87,
      "createdAt": "2026-03-15T10:30:00.000Z"
    }
  ]
}

GET/api/documents/:id/progress

Get real-time processing status for a document. Poll this endpoint after upload to track ingestion progress.

curl https://api.raas.tools/api/documents/doc_abc123/progress \
  -H "Authorization: Bearer raas_your_key"

Response

{
  "status": "processing",
  "processingStep": "embedding",
  "processingProgress": 0.72,
  "entityCount": 24,
  "chunkCount": 65
}

Search API

POST/api/search/hybrid

Combined semantic + keyword search with optional Cohere reranking. This is the recommended search endpoint for most use cases.

Request body

{
  "query": "What are the payment terms?",
  "projectId": "your-project-id",
  "limit": 5,
  "rerank": true
}

Response

{
  "data": [
    {
      "chunk": {
        "text": "Payment is due within 30 days of invoice date...",
        "metadata": { "page": 4, "section": "Terms" }
      },
      "score": 0.94,
      "document": {
        "id": "doc_abc123",
        "name": "contract.pdf"
      }
    }
  ]
}

POST/api/search/semantic

Pure vector similarity search using pgvector embeddings. Best for open-ended or conceptual queries.

{
  "query": "risks related to supply chain disruption",
  "projectId": "your-project-id",
  "limit": 10
}

GET/api/search/entities

Query extracted entities (people, dates, monetary amounts, etc.) with structured filters. Uses SQL under the hood for exact matching.

Query parameters

`projectId`	string	Project UUID (required)
`type`	string	Entity type filter (e.g. "amount", "date", "person")
`minConfidence`	number	Minimum confidence score (0-1, default 0.7)

curl "https://api.raas.tools/api/search/entities?projectId=ID&type=amount&minConfidence=0.8" \
  -H "Authorization: Bearer raas_your_key"

Response

{
  "data": [
    {
      "type": "amount",
      "value": "$3,200,000",
      "confidence": 0.95,
      "context": "Q4 Enterprise SaaS revenue totalled $3,200,000",
      "document": { "id": "doc_abc123", "name": "q4-report.pdf" }
    }
  ]
}

Chat API

POST/api/chat/conversations/:id/messages

Send a message to a document conversation. Returns an AI-generated response with source citations.

Request body

{
  "content": "What were the top revenue streams in Q4?"
}

Response

{
  "message": {
    "role": "assistant",
    "content": "The top three revenue streams in Q4 were Enterprise SaaS ($3.2M), Professional Services ($1.1M), and API Usage ($440K)."
  },
  "citations": [
    {
      "text": "Q4 Enterprise SaaS revenue totalled $3,200,000",
      "document": "q4-report.pdf",
      "page": 7
    }
  ]
}

MCP Integration

RAAS ships as a Model Context Protocol server. Connect it to Claude Desktop, Cursor, or any MCP-compatible client to give your AI tools direct access to your documents.

claude_desktop_config.json

{
  "mcpServers": {
    "raas": {
      "command": "npx",
      "args": ["-y", "@raas/mcp"],
      "env": {
        "RAAS_API_KEY": "raas_your_key",
        "RAAS_PROJECT_ID": "your-project-id"
      }
    }
  }
}

Available MCP Tools

`search_documents`	Semantic search over your document corpus
`hybrid_search`	Combined keyword + semantic search with reranking
`query_entities`	Query extracted entities (people, dates, amounts, etc.)
`compare_documents`	Analyze differences between two documents
`extract_to_format`	Extract data into a structured schema
`generate_report`	Generate a structured report from documents
`check_compliance`	Validate content against rules
`draft_from_templates`	Generate drafts using document templates

Rate Limits

Tier	API Calls	Queries/mo	Documents
Free	60/min	500	100
Pro	100/min	2,500	500
Enterprise	Custom	50,000	10,000

When rate limited, the API returns 429 with a Retry-After header.

Quick Start

A complete working example — upload a document then search it.

# 1. Upload a document
curl -X POST https://api.raas.tools/api/documents \
  -H "Authorization: Bearer raas_your_key" \
  -F "file=@contract.pdf" \
  -F "projectId=your-project-id"

# 2. Search across your documents
curl -X POST https://api.raas.tools/api/search/hybrid \
  -H "Authorization: Bearer raas_your_key" \
  -H "Content-Type: application/json" \
  -d '{
    "query": "What are the payment terms?",
    "projectId": "your-project-id",
    "limit": 5,
    "rerank": true
  }'