Understanding the technologies that power “AI”—a product-owner’s guide

There are a class of product features that make products different today than their physical or classic digital counterparts. Often they live quietly below our cursors, fingers or mouse-hovers, only throwing us off when they don’t behave as expected. This subtlety is deceptive though. Features from personalized recommendations and intelligent autofill; to user fraud detection and dynamic creative optimization have led to the massive stickiness of the modern web, and have changed the way products are built.

Having a basic understanding of the technologies and strategies that underpin these types of features, helps considerably throughout the feature development life-cycle; making it possible to design feedback loops in a way that is both flexible and leads to products that “self-improve” with more use.

In this talk I’ll discuss Condé Nast’s approach to AI-driven feature development and walk through some of the AI-driven features we’ve built across our editorial, advertising and video product portfolios. I'll also highlight the underlying infrastructure and product development habits that make these features possible.