University of Southern California

Graduate Student at USC

Los Angeles, California, United States

- Research and development of income prediction, credit default and unemployment rate models. (Monte Carlo Simulations, Logistic Regression)- Automation of rate quotes and pricing of Income Share Agreements (“ISAs”).	- End-to-End development of a "Pre-Application" that generates instant ISA rates and educates customers on the features of an ISA. Outperformed the older, manual pricing process on both business and technical metrics. (Django Web Framework, Python, Javascript, HTML, CSS, MongoDB, ReactJS)- Incorporated Google Analytics onto the "Pre-Application" to help track customer behavioural patterns and help improve user experience.- Rebuilt an updated version of the proprietary pricing model which runs 1000s of simulations in a distributed manner to generate a portfolio of cashflows and accurately price Income Share Agreements (ISAs). Improved performance by ~10x compared to the prior model. Built a REST API of the same to allow partners to price applicants in real time. (GCP, Python, Django, Spark, Monte Carlo Simulations, Regression)- Constructed pipelines to automatically push customers into select workflows on Stride's CRM (Hubspot).

Stride Funding

Software Engineer - Data Science

- Implemented high accuracy forecasting models using Box Jenkins econometrics time series modeling, producing a MAPE of below 10% for 10 of the 15 SKUs in the dataset and outperformed the existing GBM the customer had in place. (Python, Apache Spark, Google Cloud Platform, SARIMAX, Time Series Analysis using ARIMA models, Exponential Smoothing)- Developed high accuracy Convolutional Neural Networks on images for automatic asset inspections on a Windmill dataset, implementing YOLOv3 and achieving an accuracy of 70+%. (Python, YOLO v3)- Structural Equation Modeling to observe the relationship between observed variables and latent variables. (Python, SemoPy)

DDrv Analytics

Data Scientist

- Elected for the 2019-2020 school year to represent the engineering graduate student body of 6,000 students, oversee a team of 30 students, and allocate a ~$110k budget.- Served as the conduit between the student body and the administration to help raise and address concerns.

Viterbi Graduate Student Association - President

- Elected as the student representative for the graduate students in the computer science department at USC Viterbi School of Engineering. Organised social, wellness, academic and career events.

Viterbi Graduate Student Association - Informatics/Computer Science Senator

- Built, designed and rapidly brought to market new core banking products using a centralized product design tool.- Developed enhancements on the payments module of the Islamic suite of Temenos' banking product - T24.-  Participated in building a flexible and efficient award-winning Shari’ah compliant payment solution.

Varun Vegi