Abstract: AI-powered intelligent applications are growing to deliver more mainstream experiences to users. Whether predicting rainfall and its effects on specific terrain, or building a video game, AI models humanize the user experience in virtual or real world applications. Operationalizing these apps with integrated Machine learning (ML) capabilities and keeping them up to date to ensure prediction accuracy requires collaboration amongst data scientists, developers, ML engineers, IT operations, and various DevOps technologies. This can be a big lift!
In this session, we will show you how to accelerate MLOps using an example of a retail coupon application we recently built. We’ll discuss how data scientists build, test, and train ML models on Kubernetes hybrid cloud platform such as Red Hat OpenShift.
Next, we will dive into how to leverage the integrated DevOps CI/CD capabilities in Red Hat OpenShift i.e. GitOps and pipelines, to automate and accelerate the task of integrating ML models into the application development process, and ultimately deploying these applications into production at scale in a repeated way. This also helps accelerate frequent redeployment of the updated ML models into production.
Bio: Audrey Reznik has been in the IT industry (private and public sectors) for 27 years in multiple verticals. In the last 4 years, she worked as a Data Scientist at ExxonMobil where she created a Data Science Enablement team to help data scientists easily deploy ML models in a Hybrid Cloud environment. Audrey was instrumental in educating scientists about what the OpenShift platform was and how to use OpenShift containers (images) to organize, run, and visualize data analysis results. Audrey now works as a Data Scientist with the Red Hat OpenShift Data Science Team where she is focused on next-generation applications. She is passionate about Data Science and in particular the current opportunities with ML and Federated Data.