Raghav Agarwal

Master's

Data Science

Cummins Inc.

Data Scientist

• Developed an intricate Oil Reset Detection model in collaboration with Subject Matter Experts (SMEs), which included engineering a "Range Metric" using True and False positives alongside odometer data. This effort significantly improved the diagnostics pipeline, resulting in substantial cost savings and operational enhancements for Cummins. • As a Core Data Scientist, played a vital role in the Fuel Economy Optimization project, contributing to Cummins' operational efficiency and profitability with an estimated business value of $250 million. This project leveraged clustering, regression, and similarity analysis to recommend trimmable parameter combinations, establishing it as a cornerstone product for Cummins. • Contributed to a pioneering Reinforcement Learning (RL) project, creating a synthetic environment to optimize manufacturing and customer impact by dynamically adjusting prognostic model thresholds. • Successfully collaborated on a Fuel Cell Electrodes-Crack Classification Proof of Concept (POC), achieving an 88% accuracy level with limited training data. • Conducted Fuel Consumption Analysis, elucidating parameters impacting engine fuel consumption for customer insight. • Trained various models in the service sector, aiming to understand parameter correlations in the prognostics pipeline.