Each dataset is defined by a YAML semantic model. The agent is grounded in these — it will only answer using the dimensions and metrics listed here. ← back to chat
Monthly product sales by region. Synthetic data used to demonstrate the grounded agent. Revenue is in USD; units is a count.
region, product, year, month
total_revenue = SUM(revenue) [USD] — Sum of revenue (USD).total_units = SUM(units) [count] — Sum of units sold.avg_revenue = AVG(revenue) [USD] — Average revenue per record (region × product × month).2 example(s)