Serverless Machine Learning Systems with Hopsworks and Github Actions

Explore the creation of an operational prediction service using Python and free serverless services in this 31-minute Devoxx conference talk. Learn how to build CJSurf, a system that predicts wave heights for surfing at an Irish beach, updated every 6 hours. Discover how to implement feature pipelines and batch prediction pipelines using Github Actions, manage features and models with Hopsworks, and create a user interface with Streamlit. Gain insights into MLOps best practices, including automated testing, versioning, and A/B testing. Understand how Data Scientists can move beyond model training to build valuable prediction services for stakeholders using only 4 Jupyter notebooks and 1 Python program.