Benjamin Murders

Many-model training is a common use case across industries where there is a need to build hyperpersonalized models based on geography or customer. This type of training can be challenging to orchestrate and scale while maximizing resource utilization or trying to stay within budget. In this talk, learn how to leverage Snowflake ML’s distributed training capabilities to achieve the best price performance for many-model training from Container Runtime for Snowflake Notebooks. Join to hear how customer Les Schwab uses Snowflake ML to train and generate inferences on a large-series time series neural network to forecast a promotion’s success.