Let me share with you the incredible journey my team and I went on to build a machine learning-based recommendation system from scratch. It was a challenge that required the skills of Sherlock Holmes, the patience of Gandhi, and the creativity of Bob Ross. Our mission was clear: create a system that could identify substitute products in real-time, improving the online shopping experience and boosting Metro's revenue. It was a daunting task, but we were determined to succeed, armed only with our wit and an endless supply of caffeinated beverages. Do you think we succeeded? Join the talk! Let's explore the thrilling world of machine learning-based recommendation systems, talk about do's and don'ts, and maybe even learn something along the way.