Design a Demand Forecasting System — System Design Interview Practice
Design an ML-based demand forecasting system that predicts future demand across thousands of SKUs at multiple granularities, handling seasonality, promotions, and external factors. Work through the requirements, architecture trade-offs, and an interactive design review.
Requirements and concepts to consider
- Forecast demand at daily, weekly, and monthly levelsRequirement
- Handle hierarchical forecasting (store, region, national)Requirement
- Account for seasonality and holidaysRequirement
- Incorporate promotional and pricing effectsRequirement
- Include external signals (weather, economic indicators)Requirement
- Generate prediction intervalsRequirement
- Support new product forecastingRequirement
- Automate model selection per time-seriesRequirement
- mlConcept to explore
- demand forecastingConcept to explore
- time seriesConcept to explore
- supply chainConcept to explore
- retailConcept to explore
- deep learningConcept to explore