Amit Popat

Productionizing ML models doesn't have to be hard. In this talk, see how Snowflake ML makes it easy to securely and scalably deploy models for inference and manage them in production. We'll discuss how Cloudbeds chose Snowflake ML to process over 4 billion data points per hour to drive demand forecasts across 20,000 hotels with XGBoost-based models as part of its proprietary causal AI framework, resulting in 95% forecast accuracy. Cloudbeds seamlessly deploys models with Snowflake Model Registry for inference. The result is scalable causal AI forecasts that adjust room rates and drive coordinated marketing actions based on predicted demand as part of the Cloudbeds Intelligence commercial platform.