Raghav Agarwal

Professional Experience

Microsoft - Data Scientist (Intern) – Xbox Gaming

May 2024 – August 2024

  • Conducted Game Pass Impact Analysis at Xbox, identifying declines in purchases and predicting revenue loss using CausalML and Time Series models.
  • Trained predictive models to evaluate the treatment effect of Game Pass availability and analyzed counterfactual scenarios with T-learner for game releases.

George Washington University - Research Associate – Data Scientist

October 2023 – Present

  • Collaborating with the Department of Geography on multiple data science research projects, contributing to coding, data mining, analysis, and model building to advance ongoing research objectives.
  • Member of a Coast Project team as a Data Scientist, with research on Geospatial Climate Data of the Black Sea, and how changes in climate reflect on Policies, Infrastructure, and Agriculture.

Cummins Inc. - Data Scientist – Advance Analytics Team

August 2021 – August 2023

  • Successfully led a Fuel Cell Electrodes-Crack Classification POC, achieving 88% accuracy with limited training data. Secured additional funding for Manufacturing Image Analytics post-POC
  • Collaborated with SMEs on an Oil Reset Detection model, introducing 'Range Metric' with True/False positives, delivering $500K savings in Q1. Maintained a steady 25% increase in cost savings, enhancing Cummins' operational efficiency.
  • As a Core Data Scientist, contributed to the Fuel Economy Optimization project, with a business value of $250M. Used clustering, regression, and similarity analysis to optimize parameter combinations, establishing it as a key product for Cummins.
  • Contributed to a pioneering Reinforcement Learning project, creating a synthetic environment to optimize manufacturing. Adjusted prognostic model thresholds dynamically, enhancing both manufacturing efficiency and customer impact
  • Conducted Fuel Consumption Analysis to identify parameters affecting engine fuel consumption. Provided valuable insights into customer engagement and operational improvements
  • Trained models in the service sector to understand parameter correlations in the prognostics pipeline. Aimed at enhancing predictive accuracy and operational decision-making within Cummins.

Cummins Inc. - Data Engineer Intern - Advance Analytics Team

January 2021 – July 2021

  • Received one month of training in Azure Cloud, Data-Bricks, Spark, and Scala.
  • Developed data processing scripts in Scala and Spark for engine-generated data analytics.
  • Worked as a Python developer on a real-time database app, focusing on front-end and GUI using Tkinter for generative models.

Technical Skills

Awards & Honors

Education