Building Models at Scale

About the course

Building and deploying a couple models can already bring a multitude of challenges. However, scaling this up to tens or hundreds of models comes with its own unique set of challenges. In this two-day course you will learn how to build models at scale through model tracking, versioning, standardization, and automation. As well as the governance necessary to make building models at scale a success. We will dive into tracking and versioning models throughout the model lifecycle, using open-source platforms like MLflow as well as AWS and Azure services. You will learn what things should be saved and applied when building a model, which will make the whole model lifecycle more transparent . We will look into establishing audit trails which will ensure that you can always explain why and how a certain prediction was made. Additionally, you will learn how to reproduce predictions for historic models. You will also learn what parts of the model lifecycle can be standardized and automated to scale AI. For example, you will learn how to use CI/CD pipelines and how to structure repositories to (re-)train and deploy models. All the above will allow you to quickly increase the number of models running in production, while keeping tight control on the whole system.  

Why this is for you

With the constantly increasing number of models that are build and run, it also becomes increasingly important to know how to scale up your AI. This all starts with taking the right measures at the start that will facilitate this scaling. Otherwise, you might find your organization in a situation where you cannot keep track of which models are running where, what version is running and how historic predictions have been made. This poses risks to both the predictive power of the overall system as well as risks from an audit perspective.   Theory and practical use All trainings in the GAIn portfolio combine high-quality standardized training material with theory sessions from experts and hands-on experience where you directly apply the material to real-life cases. Each training is developed by top of the field practitioners which means they are full of industry examples along with practical challenges and know-how, fueling the interactive discussions during training. We believe this multi-level approach creates the ideal learning environment for participants to thrive.