Webinar "Learning through machine learning: how we built a recommendation system from scratch"

We at Metro.digital have been building data science products for years and have gathered a fair amount of experience, especially in the area of recommendation systems. Over the last few months, my team’s been focusing on an exciting new challenge: building from scratch a machine learning-based recommendation system for real-time identification of substitute products. The resulting benefits are multiple: customers quickly discover more from our assortment; Metro employees reduce their time spent on manual search; and, Metro itself increases revenue and retains customers. The aim of this talk is to tell the story of our journey from concept to production, with a focus on our learnings, mistakes, and results.