This project was developed to help BNI understand customer behavior and segmentation through accumulated data. We built an application that integrates data from multiple sources, processes large volumes through an efficient pipeline, and runs machine learning models to identify patterns and actionable insights.
The main challenge lay in infrastructure — handling data at scale while maintaining reasonable latency, and ensuring trained models remained relevant over time. We designed a modular architecture, allowing the BNI team to add features or adjust models without rebuilding the entire system.
The project remains internal and was not released to public production, but provided valuable lessons in applying machine learning within an enterprise context, managing data governance, and navigating strict compliance requirements.