Transforming Claims (40M) data to analyze usage of claims by suspicious hospitals vs. nation average.
Designing of metrics to highlight the fraudulent provider's claims easily (needed as very big dataset to analyze)
Identified and documented project constraints, assumptions, business impacts, risks and scope exclusions.
Cummins IncData Analyst Intern | May 2016 - August 2016
Fill Rate Improvement
Aditya Birla World Class ManufacturingProject Trainee | Mumbai | January 2014 - June 2014
Quality Improvement by production data analysis
Kaggle Classification competitions: Performed feature engineering on dataset and coded machine-learning models (xgboost, random forrest), Greedy Feature Selection, Ensemble multiple models, Tuned parameters with top features extraction for best score on cross-validation sets. Rank 3 in one competition
NLP: Plagiarism Detector for Gutenberg data using Locality-Sensitive Hashing (Gensim), Implemented Topic Modeling on various datasets with different statistical tools
Optimization & Simulation: Coded genetic algorithm in C for improvement decisions in transportation network of Chicago city, Improvement recommendations for an assembly line
Master of ScienceOperations ResearchTHE UNIVERSITY OF TEXAS AT AUSTIN | May 2017Relevant Coursework
Machine Learning Large Scale Data, Statistical Modeling-1, Computational Optimization, Regression
Analysis, Decision Analysis
Post-Grad DiplomaIndustrial EngineeringIIT-BOMBAY | July 2015