Data Contracts

Data Contracts

Data contracts are formal agreements between data producers and consumers — defining schema, SLOs, and data quality rules.

# Data contract YAML (OpenDataContract standard)
apiVersion: 2.2.2
kind: DataContract
id: urn:ezycoders:orders.completed:v2
info:
  title:   Orders Completed
  version: 2.0.0
  owner:   orders-team@company.com
  status:  active
  slo:
    freshness:    30 minutes
    availability: 99.9%

servers:
  production:
    type:     bigquery
    project:  myproject
    dataset:  analytics
    table:    fct_orders

models:
  - name: fct_orders
    fields:
      - name:        order_id
        type:        string
        required:    true
        unique:      true
        description: Unique order identifier
      - name:        amount
        type:        number
        minimum:     0
        maximum:     1000000
      - name:        status
        type:        string
        enum:        [completed, pending, cancelled, refunded]

# Validate contract with soda-core
# soda scan -d snowflake -c soda_config.yml orders_contract.yml

# Python validation
from datacontract.model.data_contract_specification import DataContractSpecification
contract = DataContractSpecification.from_file("orders_contract.yaml")
result   = contract.test()  # runs all quality checks