Prerequisite: None
Predictive models are often time-consuming to build and deploy, requiring expert data scientists and developers to complete effectively. Predictive models are frequently designed for a single purpose, directed at serving a marketing or product niche. Yet oftentimes, the value of predictive models can be distributed effectively across multiple functional groups with few or no modifications.
Industry expert Zank Bennett will discuss ways he has deployed single predictive models companywide to serve multiple functional groups simultaneously. He will show how, with a little bit of planning and upfront consideration, data science models can be reused, reducing costs and complexity.