NITHIN REGANTY SREENATH
Master of Science in Management Information Systems -Data Analytics (STEM Designated) Expected June 2020
University at Buffalo, The State University of New York (GPA- 3.85/4)
Relevant Courses: Database Management Systems, Statistical foundation of analytics, Predictive Analytics, Python Programming for Data Scientists, Data Visualization, Web Analytics
Bachelor’s Degree in Electronics and communication (with Distinction) Jul 2017
Visvesvaraya Technological University
Software Tools: Oracle 11g, SAS Enterprise Miner, Tableau (Certified Desktop Specialist), Visual Studio 2012, SQL Server 2012, SharePoint, MS Excel, SQLite, Amazon EC2, Amazon S3,Google Analytics (Individual Qualification)
Statistical Techniques: Regressions, Decision Trees, Random Forest, Markov Chains, clustering, Hypothesis Testing
Machine Learning: nltk, Scikit-learn, Gensim, Spacy, Keras, Natural Language Processing, Web Scraping
DATA ANALYST - CAPSTONE PROJECT Third Estate Analytics, Buffalo NY Sep 2019-Present
Pre-processed data (2.3 million records) by loading datasets to AWS EC2 storage and merged datasets using pandas.
Showcased geographical plots to the client which included parking violations and tows (2008-2019) in each street across Buffalo using Tableau and geo-pandas for data taken from Open Data Buffalo portal, which determined factors influencing the development of neighborhoods.
Built a gradient boosting regressor to rank neighborhoods based on increase in parking violations over 11 years to compare data from other similar size US cities, which helps in client’s real estate investments.
DATA ANALYST – RESEARCH PROJECT MEMBER Philanthropy Team, University at Buffalo Sep 2019-Dec 2019
Performed exploratory analysis and visualized historical data regarding alumni (2008-2018) using Tableau and showcased 10+ user interactive dashboards to provide insights to the Philanthropy Team.
Implemented a decision tree classifier and regressor using Python (Supervised Machine Learning) to classify alumni into various donor categories (Accuracy: 93%) and enhanced performance by adding a Probabilistic Model (Markov chains) to predict the probability of donation and amount of donation respectively.
DATA ANALYST – GRADUATE STUDENT ASSISTANT, University at Buffalo Sep 2019-Nov 2019
Mined data of top 200 business schools from social media platforms like Facebook, Twitter, YouTube and Instagram using web scraping tools like Selenium, Beautiful Soup libraries in Python.
Visualized and performed clustering on Tableau to see how digital reputation matters for student enrolment.
DATA SCIENCE ANALYST Accenture Solutions Pvt. Ltd., Bangalore, India Oct 2018 – May 2019
Performed EDA and regression analysis on cable modem data across different US regions to predict the likelihood of a customer calling within some time lag based on the historical pattern to create a POC for Cox Communications.
Pulled the data from AWS S3 storage using a large EC2 instance and interfaced it using boto3 library in python.
Implemented a decision tree model on sample datasets of LDPE (Low Density Polyethylene) and wine quality and built a code to extract the classification rules from output of a decision Tree.
SOFTWARE DEVELOPER Accenture Solutions Pvt. Ltd., Bangalore, India Oct 2017-Oct 2018
Developed and maintained database and User Interface of BPM portal using SQL, jQuery, Bootstrap and wrote optimized stored procedures and queries which decreased the loading time of pages by 30%.
Lead a team of 4 and revamped the Accenture Process Offerings (APO) portal to make it responsive using Bootstrap, which was not responsive before,which lead to a huge appreciation from the centre of excellence(COE) team.
DATA ANALYST INTERN, Konigtronics Pvt. Ltd., Bangalore, India Jul 2017- Oct 2017
Matched the student answers of 5th grade using answer key and reference key to generate features using NLP techniques which was fed as input to a random forest classifier to build a model to grade student essay answers.
ACADEMIC PROJECTS (GitHub_link)
CREDIT CARD FRAUD ANALYSIS USING VARIOUS SAMPLING TECHNIQUES
Balanced the highly skewed dataset which had 3 % fraud transactions using under sampling, over sampling and SMOTE techniques and built a logistic regression model to predict the fraudulent transactions (Accuracy : 99.1%).
UNDERSTANDING THE PRICING MODEL OF UBER AND LYFT THROUGH EDA AND MACHINE LEARNING
Loaded the data into SQLite database and pulled the important features (to increase efficiency) affecting prices of cab rides and compared the pricing model of the two companies to analyse which is better from a consumer standpoint.