Model serving and model metadata management are the two skills that separate the A.I. engineers above the data scientists. However there is shockingly limited tools and guides on how to do this effectively (especially compared to the effectively infinite galaxy of tutorials on LSTMs, RNNs, CNNs, and other Data Science / Machine Learning models). Kubeflow has a suite of tools that address these two areas of A.I. engineering where so few choose to tread. In this talk we'll discuss how to use Kubeflow's Artifact Store (model metadata management tool) and KFServing (Kubeflow's model serving abstraction). Users will be amazed as we quickly overview the mechanics of each piece and then dive straight into practical tips and tricks for utilizing these tools in Kubeflow.

All training courses take place 9:00am - 5:00pm, Sunday, March 15-Monday, March 16. In order to maintain a high level of hands-on learning and instructor interaction, each training course is limited in size. Participants should plan to attend both days of this 2-day training course. To attend training courses, you must register for a Platinum or Training pass; does not include access to tutorials on Monday.

In 2016, Ben Lorica, then program chair of the O'Reilly Strata Data and AI Conferences, predicted that soon "some features of AI will be incorporated into every application that we touch, and we won't be able to do anything without touching an application." For people working in data, that day has arrived. The worlds of data and AI are converging. Whether you need to learn data science with Python, build foundational data infrastructure, prepare your training data, productionize ML pipelines, work with NLP, or learn how to iterate AI projects--you'll get it all here.