Highlights –

  • Vertex AI Forecast is powerful enough to ingest datasets of up to 100 million rows from BigQuery or CSV files, covering years of historical data for thousands of product lines.
  • The tool also delivers hierarchical forecast capabilities, reducing the challenges created by organizational silos and improving overall accuracy when historical data is sparse.

With a vision to help retailers generate more accurate demand forecasts, Google has introduced a tool, Vertex AI Forecast. The new tool is part of the managed Vertex AI platform that was deployed by Google last year to help enterprises deploy machine learning faster.

Demand forecasting can have a significant impact on a retailer’s business, with elements like supply chain fluctuations and growing global market making it challenging to keep inventory stock.

Vertex AI Forecast is powerful enough to ingest datasets of up to 100 million rows from BigQuery or CSV files, covering years of historical data for thousands of product lines. The Forecast tool automatically processes the data after evaluating hundreds of different model architectures to create one model that makes managing the data relatively easier. The tool can accumulate 1,000 different demand drivers (like color, brand, promotion schedule, or e-commerce traffic statistics) and set budgets to create the forecast.

The tool also delivers hierarchical forecast capabilities, reducing the challenges created by organizational silos and improving overall accuracy when historical data is sparse. For example, the hierarchical forecast capability can bring together demand for an individual item at the store and regional levels. When the demand for individual items is too random to forecast, the model can still pick up product category-level patterns.

Google referred to a few customers already using Vertex AI Forecast, including Lowe’s, an American retail company specializing in home improvement, that used it to create accurate hierarchical models that balance SKU (Stock-Keeping Unit) and store-level forecasts.