When you pick up a product off the shelf, you may not realize the thousands of hours of work that went into developing it. R&D organizations at CPG companies are charged with developing new products that deliver real clinical or consumer benefits, are pleasant to use, have desirable aesthetics, are microbiologically robust, are shelf-stable, and that meet regulatory requirements around the world. Building new products that meet all these requirements is not simple, and for more than 100 years the work has been performed by experienced chemists using traditional scientific techniques with lots of hands-on experimentation. As CPG companies are increasingly tasked with accelerating the pace of their innovation and delivering more personalized products, it is evident that traditional approaches to product development don’t scale. This talk will focus on Colgate’s efforts to build machine learning models that learn from hundreds of thousands of historical formulations to enable scientists to digitally design new products without ever setting foot in the lab.