Optimizing Inventory for Fashion Stores using AI

Effective inventory management is critical for the dynamic fashion retail industry, particularly in men’s T- shirts. This paper presents an AI-driven solution to address overstocking and understocking, utilizing machine learning techniques to analyze product images and sales data. Key innovations includeResNet-50-based image embeddings for visual analysis, cosine similarity, and FAISS for product matching, combined with predictive modelling for demand forecasting. This approach reduces operational inefficiencies and aligns inventory levels with real-time trends, improving business profitability and sustainability.