Optiwiser A.I. Solutions

In this session, we’ll dive into how Optiwiser leverages a portfolio of 15 AI models, dynamically selected based on customer ordering behavior and enriched with secondary data such as promotions, price changes, holidays, special events (e.g. Black Friday), weather, and marketing campaigns. This multi-model approach enables us to deliver forecast precision far beyond traditional methods.

We’ll showcase the case of Oatsome, a German DTC food brand, illustrating how we designed, calibrated, and deployed the forecasting engine. The talk will highlight the key challenges we faced—such as seasonality, external factors and irregular demand spikes.

Why this talk matters / what attendees will learn:

• How to select and combine multiple forecasting models depending on SKU and order patterns
• Practical techniques for integrating external signals (promotions, events, weather, marketing) into forecasts
• Real-world lessons from deploying AI forecasting at Oatsome
• Metrics and KPIs to validate forecast accuracy and measure business impact