Schemas define the structure of the data you want to extract. Well-designed schemas produce better, more consistent results.

Schema Basics

A schema is a JSON object describing the data structure:

{
  "title": "string",
  "price": "number",
  "available": "boolean"
}

Type Reference

Primitives

{
  "name": "string",
  "count": "number",
  "active": "boolean"
}

Arrays

{
  "tags": ["string"],
  "prices": ["number"],
  "items": [{
    "name": "string",
    "qty": "number"
  }]
}

Nested Objects

{
  "product": {
    "name": "string",
    "brand": {
      "name": "string",
      "country": "string"
    }
  }
}

Best Practices

Be Specific

Instead of generic field names, use descriptive ones:

// Good
{
  "product_name": "string",
  "price_usd": "number",
  "stock_quantity": "number"
}

// Less good
{
  "name": "string",
  "price": "number",
  "qty": "number"
}

Use Appropriate Types

Match types to expected data:

{
  "rating": "number",        // 4.5
  "rating_text": "string",   // "4.5 out of 5"
  "review_count": "number"   // 1234
}

Handle Missing Data

The LLM will return null for missing fields. Design schemas to handle this gracefully.

Schema Catalog

Save and reuse schemas via the API:

# Create a schema
curl -X POST https://api.refyne.uk/api/v1/schemas \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "name": "E-commerce Product",
    "schema_yaml": "name: string\nprice: number\ndescription: string"
  }'

YAML Format

Schemas can also be defined in YAML for readability:

name: string
price: number
description: string
specifications:
  - key: string
    value: string
reviews:
  - author: string
    rating: number
    text: string

Schemas

On this page