Building a Real-Time Artificial Intelligence (AI) Platform | Machine Learning | Community Webinar

The business demand for real-time AI/ML-driven applications and use cases is gaining momentum including fraud detection, real-time product recommendations, predictive maintenance, dynamic pricing, chatbots, and more. But operationalizing AI/ML is challenging; from preparing the data for feature engineering to training models, and then deploying and monitoring them. These challenges have created a new data layer and data-transformation challenge for organizations and AI/ML professionals, including handling the proliferation and complexity of real-time feature engineering, continuous training, model serving, and monitoring. In this webinar, we will take a first-principles approach to defining a Real-Time AI Platform. Drawing from various examples across the industry, and looking at emerging architectures of some early pioneers, you will learn about some of the defining pillars of Real-Time AI Platform including feature stores, model registration, model serving, and model monitoring. Using this mental model, you will get to learn about how Redis Labs is building on the enterprise capabilities that Redis offers today, towards empowering developers and organizations to build effective and scalable Real-Time AI Platforms. About the Presenter: Taimur is a senior business executive with 18+ years of experience in new market development, strategic business development, and customer success. He has general management experience and a passion for technology, business, and public policy. Prior to Redis Labs, he led Worldwide Customer Success for Microsoft’s Azure Data Platform, Analytics & Artificial Intelligence Business. 0:00 – Introduction 8:46 – Why real-time? 11:20 – Why first principles? 12:19 – An example of operationalizing ML 13:44 – 1st first principle 23:19 – Don't forget ITBD and skills 25:09 – Early Pioneers building Feature Stores 26:34 – Doordash ML architecture 27:08 – Spotify ML architecture 30:06 – One platform, any real-time data, anywhere 31:14 – Redi