Education
-
- 2024
University at Buffalo, The State University of New York
Masters of Science: Data Science
 Core courses: Statistical Learning and Data Mining, Programming Database Fundamentals, Introduction to Machine Learning, Data Models and Query Languages, Story-telling through Data Visualization, Probability Theory, Numerical Computing and Mathematics for Data Scientists, Introduction to Computer Vision and Image Processing.
-
- 2022
Mumbai University
Bachelors of Engineering: Information Technology
 Core courses: Applied Mathematics, Data Mining and Business Intelligence, Data Structures and Algorithms, Database Management Systems, Big Data Analytics, R programming, Python Programming, Soft Computing, Internet Programming, Software Engineering and Project Management.
Experience
-
Analysis of Factors Affecting Poverty in Buffalo
Python, SQLite3, Pandas, NumPy, Seaborn
 Executed an advanced analysis to uncover factors impacting poverty in Buffalo leveraging Python, SQLite3,
Pandas, NumPy, and Seaborn.
 Managed, processed, and extracted insights from a staggering 550,000-row dataset originating from Buffalo\’s
Assessment Rolls (2017-2023).
 Unearthed 46 intricate attributes steering property value and tax dynamics within Buffalo\’s neighborhoods.
 Demonstrated analytical prowess by sculpting 6 distinct SQLite3 tables, accompanied by insightful -
ResStorage: A Recruitment Portal with Keyword Detection
Python, spaCy, BeautifulSoup4, Solidity, ReactJS, Flask
 Conceptualized and built ResStorage: A ReactJS and Flask-powered web app combining NLP-based resume
keyword highlighting, scoring, Ethereum-based document storage, and LinkedIn data verification.
 Designed an NLP model using spaCy to identify job-relevant keywords in resumes.
 Implemented a scoring system leveraging Supervised Learning (Decision Tree classification) for effective resume
evaluation.
 Developed secure Ethereum-based document storage with Solidity for confidential files deploying MetaMask for
transactions.
 Utilized web scraping with BeautifulSoup and Selenium to improve candidate profile verification through LinkedIn.
 Collected candidate data from LinkedIn to further verify candidate profile using web-scraping. -
Multiclass Digit Recognition using CNN on SVHN Dataset
Python, Pytorch
 Modeled a CNN (AlexNet) employing Pytorch to accurately perform digit classification on the SVHN dataset,
consisting of 10-digit classes with 73,257 training and 26,032 testing digits.
 Performed comprehensive image preprocessing, including resizing, tensor conversion, normalization, and
efficient data loader implementation deploying Pytorch for streamlined model training.
 Engineered an adaptable AlexNet architecture in PyTorch, optimizing hyperparameters, achieving a remarkable
model performance with a test accuracy of 93.93%.
 Achieved a robust test accuracy of 91.77% by retraining the model with the augmented dataset, leveraging the
optimized setup. -
Fantasy Premier League Analysis
PostgreSQL, Microsoft Excel, Python
 Employed PostgreSQL, Excel, and Python to dissect 6-season Fantasy Premier League data.
 Structured it ingeniously with normalized tables, each meticulously brought to BCNF form to foster seamless and
efficient analysis.
 Boosted query performance by strategically deploying B-tree indexing on the pivotal player_id attribute.
 Designed an intuitive Streamlit interface, enhancing the experience for end users.