This technical blog outlines how we built a scalable inventory optimization system to help partners maintain a profitable inventory.

Zalando Tech Blog is a platform where Zalando's engineering team shares insights, experiences, and best practices on technology and innovation. Covering topics such as software architecture, data science, and machine learning, Zalando's Tech Blog offers resources for developers and tech enthusiasts. Developers can learn about Zalando's engineering culture, projects, and technology stack through their blog posts and articles.

Zalando

Zalando built a scalable inventory optimization system that combines demand forecasting and replenishment optimization to help e-commerce partners manage millions of SKUs across multiple warehouses. The system uses a two-step approach: generating probabilistic demand forecasts with LightGBM and Nixtla's MLForecast, then applying Monte Carlo simulations with gradient-free optimizers to minimize inventory costs. The architecture leverages AWS services including SageMaker, Databricks, and feature stores to provide both real-time interactive recommendations and daily batch processing, handling 5 million SKUs with 2-hour processing times.

Building a dynamic inventory optimisation system: A deep dive

Overarching building blocks and design philosophy

Scalable demand forecasts for millions of articles

Translating Demand Forecasts into Actionable Inventory Strategies