• Spearhead the development of cross-vertical machine learning toolkits for internal data scientists to analyze fuel usage data from IoT and telematics devices in the fleet, rail and heavy equipment/machinery industries.• Create and maintain codebase to infer operator behaviors from sensor data in the fleet industry, such as braking, acceleration, cornering and excess idling.• Reduced fleet customer’s engine idle time by 40%, which translated to an estimated $4 million in annual savings.• Collaborate closely with product and engineering leaders to ideate and create best-in-class software features to track and predict fuel usage.

• Developed models to estimate the remaining useful life of components in industrial machinery with high confidence and accuracy, leading to a reduction in unplanned down time and service costs for customers.• Built reporting tools and the infrastructure to ascertain the impact of our models to our customers’ bottom line.• Worked closely with developers to build model deployment tools and data processing pipelines.• Used R Shiny, flexdashboard and htmlwidgets to build interactive dashboards and data visualizations.

• Provide custom data mining and analytics solutions for clients.• Pitch data science projects and build demonstration models for prospective clients.Projects:NLP Product Summary -- automatically scrape product reviews from e-commerce websites and extract customers' opinions on specific features of a product.Customer Segmentation -- segment a client's customers based on past purchases and "firmographic" information.

• Designed and optimized hospital pharmacy call center to better meet demand requirements and improve customer service.• Conducted statistical experiments to determine the probability distributions of incoming calls as well as service times.• Determined optimum staffing levels using the concepts of queuing theory and Markov chain methods. • Published paper about findings in a healthcare optimization journal.