deepread

DeepRead - Production OCR API

DeepRead is an AI-native OCR platform that turns documents into high-accuracy data in minutes. Using multi-model consensus, DeepRead achieves 97%+ accuracy and flags only uncertain fields for Human-in-the-Loop (HIL) review—reducing manual work from 100% to 5-10%. Zero prompt engineering required.

What This Skill Does

DeepRead is a production-grade document processing API that gives you high-accuracy structured data output in minutes with human review flagging so manual review is limited to the flagged exceptions

Core Features:

• Text Extraction: Convert PDFs and images to clean markdown
• Structured Data: Extract JSON fields with confidence scores
• HIL Interface: Built-in Human-in-the-Loop review — uncertain fields are flagged (hil_flag) so only exceptions need manual review
• Multi-Pass Processing: Multiple validation passes for maximum accuracy
• Multi-Model Consensus: Cross-validation between models for reliability
• Free Tier: 2,000 pages/month (no credit card required)

Setup

1. Get Your API Key

Sign up and create an API key:

# Visit the dashboard
https://www.deepread.tech/dashboard

# Or use this direct link
https://www.deepread.tech/dashboard/?utm_source=clawdhub

Save your API key:

export DEEPREAD_API_KEY="sk_live_your_key_here"

2. Clawdbot Configuration (Optional)

Add to your clawdbot.config.json5:

{
  skills: {
    entries: {
      "deepread": {
        enabled: true
        // API key is read from DEEPREAD_API_KEY environment variable
        // Do NOT hardcode your API key here
      }
    }
  }
}

3. Process Your First Document

Option A: With Webhook (Recommended)

# Upload PDF with webhook notification
curl -X POST https://api.deepread.tech/v1/process \
  -H "X-API-Key: $DEEPREAD_API_KEY" \
  -F "[email protected]" \
  -F "webhook_url=https://your-app.com/webhooks/deepread"

# Returns immediately
{
  "id": "550e8400-e29b-41d4-a716-446655440000",
  "status": "queued"
}

# Your webhook receives results when processing completes (2-5 minutes)

Option B: Poll for Results

# Upload PDF without webhook
curl -X POST https://api.deepread.tech/v1/process \
  -H "X-API-Key: $DEEPREAD_API_KEY" \
  -F "[email protected]"

# Returns immediately
{
  "id": "550e8400-e29b-41d4-a716-446655440000",
  "status": "queued"
}

# Poll until completed
curl https://api.deepread.tech/v1/jobs/550e8400-e29b-41d4-a716-446655440000 \
  -H "X-API-Key: $DEEPREAD_API_KEY"

Usage Examples

Basic OCR (Text Only)

Extract text as clean markdown:

# With webhook (recommended)
curl -X POST https://api.deepread.tech/v1/process \
  -H "X-API-Key: $DEEPREAD_API_KEY" \
  -F "[email protected]" \
  -F "webhook_url=https://your-app.com/webhook"

# OR poll for completion
curl -X POST https://api.deepread.tech/v1/process \
  -H "X-API-Key: $DEEPREAD_API_KEY" \
  -F "[email protected]"

# Then poll
curl https://api.deepread.tech/v1/jobs/JOB_ID \
  -H "X-API-Key: $DEEPREAD_API_KEY"

Response when completed:

{
  "id": "550e8400-...",
  "status": "completed",
  "result": {
    "text": "# INVOICE\n\n**Vendor:** Acme Corp\n**Total:** $1,250.00..."
  }
}

Structured Data Extraction

Extract specific fields with confidence scoring:

curl -X POST https://api.deepread.tech/v1/process \
  -H "X-API-Key: $DEEPREAD_API_KEY" \
  -F "[email protected]" \
  -F 'schema={
    "type": "object",
    "properties": {
      "vendor": {
        "type": "string",
        "description": "Vendor company name"
      },
      "total": {
        "type": "number",
        "description": "Total invoice amount"
      },
      "invoice_date": {
        "type": "string",
        "description": "Invoice date in MM/DD/YYYY format"
      }
    }
  }'

Response includes confidence flags:

{
  "status": "completed",
  "result": {
    "text": "# INVOICE\n\n**Vendor:** Acme Corp...",
    "data": {
      "vendor": {
        "value": "Acme Corp",
        "hil_flag": false,
        "found_on_page": 1
      },
      "total": {
        "value": 1250.00,
        "hil_flag": false,
        "found_on_page": 1
      },
      "invoice_date": {
        "value": "2024-10-??",
        "hil_flag": true,
        "reason": "Date partially obscured",
        "found_on_page": 1
      }
    },
    "metadata": {
      "fields_requiring_review": 1,
      "total_fields": 3,
      "review_percentage": 33.3
    }
  }
}

Complex Schemas (Nested Data)

Extract arrays and nested objects:

curl -X POST https://api.deepread.tech/v1/process \
  -H "X-API-Key: $DEEPREAD_API_KEY" \
  -F "[email protected]" \
  -F 'schema={
    "type": "object",
    "properties": {
      "vendor": {"type": "string"},
      "total": {"type": "number"},
      "line_items": {
        "type": "array",
        "items": {
          "type": "object",
          "properties": {
            "description": {"type": "string"},
            "quantity": {"type": "number"},
            "price": {"type": "number"}
          }
        }
      }
    }
  }'

Page-by-Page Breakdown

Get per-page OCR results with quality flags:

curl -X POST https://api.deepread.tech/v1/process \
  -H "X-API-Key: $DEEPREAD_API_KEY" \
  -F "[email protected]" \
  -F "include_pages=true"

Response:

{
  "result": {
    "text": "Combined text from all pages...",
    "pages": [
      {
        "page_number": 1,
        "text": "# Contract Agreement\n\n...",
        "hil_flag": false
      },
      {
        "page_number": 2,
        "text": "Terms and C??diti??s...",
        "hil_flag": true,
        "reason": "Multiple unrecognized characters"
      }
    ],
    "metadata": {
      "pages_requiring_review": 1,
      "total_pages": 2
      }
  }
}

When to Use This Skill

✅ Use DeepRead For:

• Invoice Processing: Extract vendor, totals, line items
• Receipt OCR: Parse merchant, items, totals
• Contract Analysis: Extract parties, dates, terms
• Form Digitization: Convert paper forms to structured data
• Document Workflows: Any process requiring OCR + data extraction
• Quality-Critical Apps: When you need to know which extractions are uncertain

❌ Don't Use For:

• Real-time Processing: Processing takes 2-5 minutes (async workflow)
• Batch >2,000 pages/month: Upgrade to PRO or SCALE tier

How It Works

Multi-Pass Pipeline

PDF → Convert → Rotate Correction → OCR → Multi-Model Validation → Extract → Done

The pipeline automatically handles:

• Document rotation and orientation correction
• Multi-pass validation for accuracy
• Cross-model consensus for reliability
• Field-level confidence scoring

Human-in-the-Loop (HIL) Interface

DeepRead includes a built-in Human-in-the-Loop (HIL) review system. The AI compares extracted text to the original image and sets hil_flag on each field:

• hil_flag: false = Clear, confident extraction → Auto-process
• hil_flag: true = Uncertain extraction → Routed to human review

How HIL works:

1. Fields extracted with high confidence are auto-approved
2. Uncertain fields are flagged with hil_flag: true and a reason
3. Only flagged fields need human review (typically 5-10% of total fields)
4. Review flagged fields in DeepRead Preview (preview.deepread.tech) — a dedicated HIL review interface where reviewers can see the original document side-by-side with extracted data, correct flagged fields, and approve results
5. Or integrate with your own review queue using the hil_flag data in the API response

AI flags extractions when:

• Text is handwritten, blurry, or low quality
• Multiple possible interpretations exist
• Characters are partially visible or unclear
• Field not found in document

This is multimodal AI determination, not rule-based.

Advanced Features

1. Blueprints (Optimized Schemas)

Create reusable, optimized schemas for specific document types:

# List your blueprints
curl https://api.deepread.tech/v1/blueprints \
  -H "X-API-Key: $DEEPREAD_API_KEY"

# Use blueprint instead of inline schema
curl -X POST https://api.deepread.tech/v1/process \
  -H "X-API-Key: $DEEPREAD_API_KEY" \
  -F "[email protected]" \
  -F "blueprint_id=660e8400-e29b-41d4-a716-446655440001"

Benefits:

• 20-30% accuracy improvement over baseline schemas
• Reusable across similar documents
• Versioned with rollback support

How to create blueprints:

# Create a blueprint from training data
curl -X POST https://api.deepread.tech/v1/optimize \
  -H "X-API-Key: $DEEPREAD_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "name": "utility_invoice",
    "description": "Optimized for utility invoices",
    "document_type": "invoice",
    "initial_schema": {
      "type": "object",
      "properties": {
        "vendor": {"type": "string", "description": "Vendor name"},
        "total": {"type": "number", "description": "Total amount"}
      }
    },
    "training_documents": ["doc1.pdf", "doc2.pdf", "doc3.pdf"],
    "ground_truth_data": [
      {"vendor": "Acme Power", "total": 125.50},
      {"vendor": "City Electric", "total": 89.25}
    ],
    "target_accuracy": 95.0,
    "max_iterations": 5
  }'

# Returns: {"job_id": "...", "blueprint_id": "...", "status": "pending"}

# Check optimization status
curl https://api.deepread.tech/v1/blueprints/jobs/JOB_ID \
  -H "X-API-Key: $DEEPREAD_API_KEY"

# Use blueprint (once completed)
curl -X POST https://api.deepread.tech/v1/process \
  -H "X-API-Key: $DEEPREAD_API_KEY" \
  -F "[email protected]" \
  -F "blueprint_id=BLUEPRINT_ID"

2. Webhooks (Recommended for Production)

Get notified when processing completes instead of polling:

curl -X POST https://api.deepread.tech/v1/process \
  -H "X-API-Key: $DEEPREAD_API_KEY" \
  -F "[email protected]" \
  -F "webhook_url=https://your-app.com/webhooks/deepre

---

*Content truncated.*