Skip to main content
Transform raw product data into complete, structured catalogs. Every field is validated, sourced, and ready for production. Fully compatible with your existing PIM and ERP schemas.

Quickstart

Enrich your first item in 2 minutes

API Reference

Full endpoint documentation

​
Why Rastro?

​
Auto-Scaling Infrastructure

Processing 30,000+ SKUs is complex — managing queues, handling failures, rate limiting, and scaling workers. Rastro handles all of this for you. Just send your data and we process it at scale.

​
Web or Internal Sources with Full Traceability

Rastro can fetch information for you from anywhere on the internet with no URLs provided, PDFs, images, and any data or links you provide. Every enriched value includes sources showing where data was found, plus optional explanations of how values were derived. No black boxes—audit any data point back to its origin: 
{
  "after_data": {
    "material": "316 Stainless Steel",
    "bore_diameter": "25 mm",
    "datasheet_url": "https://example.com/spec.pdf",
    "sources": {
      "material": ["https://mcmaster.com/catalog/91251A"],
      "bore_diameter": ["https://mcmaster.com/catalog/91251A"],
      "datasheet_url": ["https://mcmaster.com/catalog/91251A"]
    },
    "source_explanations": {
      "material": "Extracted from product title and specifications table",
      "bore_diameter": "Extracted from product dimensions section",
      "datasheet_url": "Extracted from specifications table on page 2"
    }
  },
  "all_sources": ["https://mcmaster.com/catalog/91251A"]
}
Set "web_search": false to skip web lookups and only process your input data. Sources will reference INPUT_DATA instead of URLs.

​
Automatic Error Flagging

AI flags uncertain values, conflicting data, and potential issues for human review. Every result includes reasoning and flags. 
{
  "review_info": {
    "reasoning": "Found conflicting specs across sources",
    "flags": ["conflicting_data", "verify_dimensions"],
    "flag_record": true
  }
}

​
Taxonomy & Category Prediction

Automatically classify items into your taxonomy with attribute extraction. Define your category hierarchy once—Rastro assigns items and extracts category-specific attributes. 
{
  "category_path": "Fasteners > Bolts > Hex Bolts",
  "taxonomy_attributes": {
    "Thread Size": "M10-1.5",
    "Head Type": "Hex",
    "Grade": "A2-70"
  }
}

​
Complex Field Types

Define fields with units, enums, arrays, and validation constraints. Rastro extracts structured data matching your exact schema. 
{
  "output_schema": [
    {"name": "weight", "type": "number", "unit": "kg"},
    {"name": "material", "type": "string", "enum": ["Steel", "Aluminum", "Brass"]},
    {"name": "certifications", "type": "array", "items_enum": ["ISO 9001", "CE", "UL"]}
  ]
}

​
Quality Scoring

Get a 1-5 readiness score for each item based on your criteria. Identify incomplete records before they hit production. 
{
  "quality_score": 4,
  "quality_result": {
    "explanation": "Complete specs, missing datasheet URL",
    "issues": ["No datasheet link"],
    "suggestions": ["Add manufacturer datasheet"]
  }
}

Enrich API

Web research with citations

Flows

Visual data pipelines

Catalogs

Managed product storage

MCP (Codex / Claude)

Agent-driven bulk catalog operations with review safety