Time Series Forecasting of Programmatic Inventory at Scale

Visualizations/ UI/ Business Use Cases

When advertisers connect with GumGum with their targeting details, it is important to know if we have enough matching inventory from the publisher network to serve the ads. At GumGum, we have ~10B programmatic inventory impressions per day. The goal is to enable the business to make real time predictions about the volume of inventory using time series forecasting models at this huge scale.