EDUCATION
-
Masters in Data Science, Analytics and Engineering (Computing and Decision Analytics) |
December 2025 |
Arizona State University (GPA 3.78/4) |
Tempe, USA |
Coursework: Data Mining, Statistical Machine Learning and Statistics for Data Analysts
-
Bachelor of Technology in Mechanical Engineering with Honors |
July 2022 |
SASTRA University (GPA 3.6/4) |
Thanjavur, India |
Coursework: Programming in C, Object-Oriented Programming in C++, Engineering Mathematics, Statistical Methods and Resource Management
TECHNICAL SKILLS
- Programming: Python (NumPy, Pandas, scikit-learn, TensorFlow, PyTorch, spaCy, NLTK), Java, SQL (MySQL, PostgreSQL).
- Data Wrangling: Web scraping (BeautifulSoup, Selenium), data cleaning, preprocessing, and transformation.
- Visualization and Analysis: Tableau, Microsoft Excel, Exploratory Data Analysis (EDA), regression models, and time series analysis.
- Statistical Techniques: Hypothesis Testing, Bayesian Inference, Bootstrapping, and Statistical Inference.
PROJECT EXPERIENCE
-
Energy consumption forecasting |
January 2024 – May 2024 |
- Engineered a time-series predictive model using XGBoost and deep learning (LSTM), achieving a 15% improvement in accuracy.
- Identified $223,110 in potential annual savings through quantitative analysis of energy inefficiencies across 24 buildings, contributing to 30% of energy costs.
- These findings supported strategic decision-making in energy optimization, reducing operational costs.
- Related Resources
-
Spotify Music Popularity Analysis |
October 2024 – December 2024 |
- Analyzed Spotify data using feature engineering, statistical tests, and ML models (84.3% accuracy) to predict popularity and classify moods.
- Identified 5 emerging artists with a 30% stream growth potential and optimized playlists for 15% higher engagement.
- Insights facilitated improved user engagement strategies for the Spotify platform.
- Related Resources
-
Text and Sentiment Analysis of Political Rally Speeches |
May 2024 – June 2024 |
- Executed sentiment analysis on over 30 rally speeches (18,000 tokens each), leveraging complex data sets and quantitative analysis using Python (NLTK, SpaCy).
- Applied data visualization techniques to extract sentiment trends and word frequencies, optimizing summarization using BART and parallel processing to reduce summarization time by 30%.
- Findings contributed to better understanding of public sentiment trends in political contexts.
- Related Resources
-
Optimizing Credit Card Fraud Detection |
January 2024 – May 2024 |
- Designed and built a credit card fraud detection system using quantitative analysis on 280,000 transactions, applying modeling techniques (RandomForest, neural networks, logistic regression).
- Improved detection accuracy by 20% and decreased false positives by 15% through data manipulation and feature engineering with Python.
- Results supported fraud mitigation strategies, reducing financial risks for credit providers.
- Related Resources
-
Sentiment Analysis with DistilBERT on App reviews |
May 2024 – July 2024 |
- Constructed and implemented a sentiment analysis model to evaluate over 6,000 user reviews from five productivity apps by deploying the Google Play Scraper API.
- Optimized preprocessing for efficient CPU performance and implemented the model using PyTorch, achieving a 15% improvement in sentiment classification accuracy and identifying key app features impacting user satisfaction.
- Related Resources
PROFESSIONAL EXPERIENCE