Skip to main content

Example workflows

Concrete schemas you can map onto Lake → gold → train. No fake metrics.

Churn (classification)​

Gold columns: customer_id, days_since_login, support_tickets, plan_tier, churned (target)

  1. Upload CSV to bronze
  2. Publish gold with use case Machine learning, target churned
  3. Train → deploy

Demand (forecast)​

Gold columns: order_date (datetime), units_sold (value)

  1. Upload history to bronze
  2. Publish gold with use case Forecast
  3. Build a forecast

Image classes (unstructured → ML)​

Zip layout: one root folder, then class folders with .jpg / .png (layout rules)

  1. Unstructured zip ingest → bronze
  2. Publish gold for ML (target = label if present)
  3. Train in Machine Learning

Next​