This paper provides a comprehensive overview of a machine learning based model developed for the collaborative logistics and finance sectors. At the core of the model is the Optimisation Vehicle Routing Problem (VRP), which improves logistics efficiency by scientifically allocating delivery vehicles and planning routes. The model leverages the rich dataset of the e-commerce XB platform to integrate consumer behaviour and demand patterns into its predictive analytics framework. Key to this approach is the implementation of feature engineering, which involves data preprocessing, transformation and selection to ensure that valuable insights are extracted and predictive accuracy is improved. In addition, the model employs an improved back-propagation neural network algorithm, which is essential for reducing errors and refining forecasting performance in complex logistics environments. The inclusion of a grey GM(1,1) model for forecasting logistics demand further enhances the model’s ability to predict freight volume trends and help make strategic decisions. This development not only demonstrates the potential of machine learning to transform logistics operations and financial planning, but also highlights the need for continuous adaptation and refinement in response to dynamic market and consumer behaviour. This collaborative model demonstrates the power of combining advanced computing technologies with practical logistics strategies, providing a promising avenue for future growth in these interconnected fields.