Transforming the Retail Industry with Transformers

Explore the challenges and opportunities of implementing transformer models and pretrained language models in the retail industry through this 40-minute conference talk from the Toronto Machine Learning Series. Discover how SPINS, a data/tech company focused on the natural grocery industry, leverages state-of-the-art language models to automate data ingestion workflows and drive business outcomes. Learn about the end-to-end MLOps system used to mitigate challenges associated with deploying complex AI models, particularly for small and mid-sized companies with limited resources. Gain insights on when to choose massive transformer models over classical ML models, based on real-world experimentation with ill-defined label spaces, numerous classes, and data imbalance. Apply the speakers' findings and recommendations to solve problems more effectively and efficiently in various real-world settings within the retail industry and beyond.