Cloud Meetup 012: Berkeley Changemaker, iSchool, and Cricket Health (03/31/2020)

Kunal is a data scientist working at Cricket Health, a specialty kidney care company, targeting chronic kidney disease (CKD). CKD is a massive problem, with over $100 billion in annual direct healthcare costs to Medicare alone, and has very little visibility and few symptoms prior to kidney failure — a silent killer. As a data scientist, he builds and deploys first-of-their-kind predictive models based on claims data for screening undiagnosed CKD and hospitalization risk, working with and training ML models on very large payer datasets in a HIPAA compliant way. In his discussion, he’ll talk through approaches Cricket has taken in the 3 distinct phases of modeling: 1) outcome engineering, 2) feature engineering (via a predictive pipeline), and 3) training in the cloud.