CASE Studies
Phase 1: Product Scope
Reduced product risk and increased revenue streams for therapeutic device start-up by advising founders to scope down offering from smart apparel to HaaS
Start-up advisor + mentor to 10+ founders on development and implementation of strategies to evaluate and improve product/market fit based on customer and product understanding, from idea to MVP
Phase 2: Data Strategy
Advised CEO and Engineering team of hospitality start-up on how to close gaps in company’s data collection pipeline for tracking customer service requests; quarterly advising and code reviewing analytics initiatives in Python
Spearheaded efforts to standardize data quality workflows for tracking quality of care metrics across two disparate Electronic Health Record (EHR) systems of a major healthcare organization
Phase 3: Business Metrics
Developed positive and negative KPIs to evaluate quality of network connectivity for online gaming start-up with terabytes of data (220M+ requests/day)
Prioritized 40+ initiatives in collaboration with executives of major entertainment studio, based on product payoff, timeline to delivery and data availability
Collaborated with CEO and Chief Medical Officer to evaluate and improve existing healthcare quality metrics, sharing findings with physicians and nurses of 100+ US hospitals
Phase 4: Product Improvement
Identified areas of missed revenue opportunities, by understanding nuances of how customers choose what to buy, for fashion subscription box from 10K+ customer reviews
Built end-to-end near-real time recommendation engine for a social media start-up to improve engagement of social media posts via ExtraTrees, and productionalized as an API via AWS SageMaker, Python and Docker
Minimized oil rig equipment failure and future potential crises by predicting when oil rig needs maintenance, from IoT sensors for manufacturing company via Elastic Nets in Python on AWS EC2
Increased GPS accuracy for point-of-interest detection system in IoT mobile devices by 15% for mobile market research start-up by leveraging Google API and Boosted Trees in Python
Invented a Novel Feature Engineering Approach for Healthcare Data for unequally-spaced, age-dependent patient vitals, for predicting out-of-sample ICU outcomes with 8% more accuracy; presented at MUCMD
Generated $4M+ annually across 1.5K+ global studies and multiple CPG clients, by leading development of an end-to-end R/shiny app, for estimating causal structure and strength of key drivers of CPG purchases