ZEOS Builds Dynamic Inventory Optimization System
ZEOS is developing an AI-driven dynamic inventory optimisation system to provide replenishment recommendations for e-commerce partners, targeting millions of articles and using a 2.5-year history with weekly probabilistic forecasts. The system combines LightGBM via Nixtla MLForecast, Monte Carlo simulations, gradient-free optimisers, and zFlow infrastructure on AWS/Databricks to deliver daily batch and real-time endpoints for partners, reducing stockouts and holding costs.
Scoring Rationale
Practical, production-grade architecture and forecasting details drive score; limited research novelty and quantitative outcomes constrain higher impact.
Practice with real Retail & eCommerce data
90 SQL & Python problems · 15 industry datasets
250 free problems · No credit card
See all Retail & eCommerce problems
