Abstract: While some problems can be solved with a single, end-to-end "mega" model that go from raw inputs to end outputs, practitioners (including our customers!) find that these approaches don’t replace the engineering advantages of a modular, composable pipeline of individual building blocks— including custom models, preprocessing steps, and business logic. Like any other piece of production infrastructure or code, it’s critical for practitioners to inspect, test, modify, and swap the modular components in an end-to-end system. In this talk, we'll discuss a data-centric approach to development, which anchors on modular components, fine-grained evaluation, and adaptability through programmatic labeling. Along the way, we'll share a real-world case study of an AI application through these principles are implemented to drive business and product value.
Bio: Bio Coming Soon!
Vincent Sunn Chen
Founding Engineer, Lead ML Engineer | Snorkel AI