ProcessMiner Inc.

• Implemented a time series AI solution for a leading US based paper manufacturer to predict the quality of the paper that is under production, beforehand, with an accuracy of 80% based on real-time streaming sensor data (R, SQL)
• Designed and implemented a novel random optimization algorithm for balanced hyper-parameter tuning using space-filling curves and design of experiments resulting in 20% increase in accuracy of the prediction system (Python, SQL)
• Engineered the creation of new features for the classification system resulting in a 25% increase in prediction performance
• Developed an open-source package ‘Timeoracle’ using state-of-the-art techniques for Time Series Modelling and Forecasting that is set to have a public release in 2021 (Python)
• Built a prototype for an AutoML (Automated Machine Learning) framework to increase the company’s AI footprint and speed up on-boarding of new customers slashing time efforts from 2 weeks to 8 hours (Python, SQL)
• Developed a novel root-cause system to identify process variables that led to the production of a defective unit using PCA and Adaptive Lasso for a high dimensional real-time data stream reducing defect rate by 40%(Python, SQL)
• Implemented the root cause system for 5 paper mills and presented the analysis report to stakeholders leading to advanced visibility on the paper manufacturing process and contributing to instant 15% decrease in weekly maintenance costs

Infosys

• Actioned transactional data analysis from various banks and performed extensive feature engineering to improve the money laundering pattern detection system’s accuracy by 15% (R, SQL)
• Forecasted the weekly defect rate by building a platform to extract, transform data and train Random Forest, Gradient Boosting models aiding in effective decision-making on product release dates (R, SQL)
• Authored 5+ business intelligence reports providing advanced visibility in predicting daily, weekly and monthly operational performance of the team using forecasting models (Tableau, Excel)
• Performed RFM analysis to identify customer purchase patterns to create segments and to analyze customer behavior across segments to boost sales of Cisco products by at least 15% for each major customer segment (Python, SQL)
• Optimized data fetch time from ~14 mins to 0.5 secs by creating indices, revamping complex SQL queries and schema structure

Infosys

• Spearheaded a team of 8 in developing a cloud solution that helped customers install 5+ products on a remote server with zero customer interaction contributing $15M+ earnings to the client (NodeJS, MongoDB)
• Resolved several critical high-impact issues, defects for important stakeholders improving customer satisfaction scores by 15%

The University of Texas at Dallas

GPA: 3.9

• Recipient of Dean's Excellence scholarship for 4 continuous terms
• Served as Graduate Teaching Assistant for courses: Business Analytics with R, Advanced Business Analytics with R, Database Systems etc.
• Participated in several Hackathons and case competitions proving my analytics, machine learning skills
• Served as the Operations Lead at INFORMS UTD

Amrita School of Engineering, Coimbatore, India

GPA: 3.7

• Graduated with a Distinction (First Class)
• Soccer player. Represented the university at various touranments