Serving ML models is one of the most complex steps when it comes to AI/ML in production, as you have to put all the pieces together into a unified system while considering:

• throughput/latency requirements

• infrastructure costs

• data and model access

• training-serving skew

.

As we started this project with production in mind by using the Hopsworks AI Lakehouse, we can easily bypass most of these issues, such as:

• the query and ranking models are accessed from the model registry;

• the customer and H&M article features are accessed from the feature store using the offline and online stores depending on throughput/latency requirements;

• the features are accessed from a single source of truth (feature store), solving the training-serving skew.

.

Estimating infrastructure costs in a PoC is more complicated. Still, we will leverage a Kubernetes cluster managed by Hopsworks, which uses KServe to scale up and down our real-time personalized recommender depending on traffic.

.

Thus, in this lesson, you will learn how to:

• Architect offline and online inference pipelines using MLOps best practices.

• Implement offline and online pipelines for an H&M real-time personalized recommender.

• Deploy the online inference pipeline using the KServe engine.

• Test the H&M personalized recommender from a Streamlit app.

• Deploy the offline ML pipelines using GitHub Actions.

.

Enjoy!