From Data to Production: An End-to-End Automated ML Pipeline in Snowflake

Many companies have data in databases that they'd love to use for ML workloads, but the field is so daunting and the amount of technology to understand is so massive that it can sometimes feel impossible to tackle.

This HoL, as the final lab in the three part Pipeline series, aims to resolve this issue by giving participants a step-by-step framework for setting up automated, replicable pipelines for their ML workloads. After setup, these pipelines will be capable of running and reporting as autonomously as participants desire, all in one place under a single governance framework.

In this Hands-on Lab, we will be diving into automating as much of the standard machine learning pipeline as we can, be it ensuring data freshness, retraining models, comparing metrics, and monitoring model drift.

By the end of this Hands-on Lab, participants will gain a replicable toolkit they can use to deploy every feature and automation setup discussed in this lab in their own environments, as well as an understanding of how easy these features and automations are to implement with Snowflake's infrastructure and how each part works behind the scenes.