Data Science, Machine LEarning

Venkatasai Varada

*** **** ****** • Hartford, CT • 860-***-****• **********.******@*****.***

venkatasai-varada.strikingly.com•linkedin.com/in/venkatasaivarada •github.com/venkatsaivvs

SAS Certified Statistical Business Analyst

EDUCATION:

The University of Connecticut, School of Business Hartford, CT

Master of Science in Business Analytics and Project Management, GPA: 3.8/4.0 Expected May 2017

Jawaharlal Nehru Technological University Hyderabad, India

Bachelor of Technology in Electronics and Instrument Engineering, Percentage: 82.8/100.0 April 2014

SKILLS:

Python, R, SQL, PL SQL, Azure ml, Google analytics, A/B testing, UNIX (Shell Scripting), Tableau, SAS: Base, JMP, Enterprise miner, Hadoop: Pig, Hive, Text analytics, Time series forecasting: ARIMA models, Machine Learning: Regression, GLM, Decision tree, Random Forest, Boosting, Bagging, SVM, Neural networks, NLP, Deep Learning: MxNet, deepr;

EXPERIENCE:

Sykes Enterprises Inc December 2016 – Present

Data Scientist Intern

Perform Social network analysis on chat data to understand how social networking can affect attrition, absenteeism and adaptability to business. Analyze betweenness, degree, closeness centralities of a social network to understand influencers

Perform text analytics and topic modeling to analyze chat data of customer representatives in order to identify frequent help seekers. Also, understand topics of discussion and major issues that representatives find difficult to answer thereby providing specialty training to efficiently answers calls

Built a predictive model to alert supervisors that agents may need help based on factors including call duration, clicks data so that assistance can be provided to agents hesitating to ask help

University of Connecticut, Graduate Consulting Club, Hartford, CT January 2016-April 2016

Graduate Consultant: BiCi (Bicycle Community) Co.

Predicted whether a customer would renew subscription or not, so that the campaign team would target right clients to maximize number of renewal subscriptions and understand customer behavior towards joining Bicycle community. Performed logistic regression on customer data to build the model

Built database to store membership data and automated download and import into database processes by writing python scripts to eliminate manual intervention by 70% and paper based record management system by 100%

TATA Consultancy Services, Hyderabad, India June 2014-December 2015

Assistant System Engineer (Multi Channel Customer Interface)

Developed various web applications to provide online services including monitoring mobile data usage, online recharge, bill payments for telecom customers which reduced customer calls to the representatives by 50%

Automated manual cash collection and reconciliation process by developing collection accounting and automated systems through shell scripts which eliminated manual interaction by 85% and saved 3 workhours/day

Facilitated host-to-host solution with bank ensuring complete security of data and elimination of manual exchange of information between banker and client, as well as delay in collections debited to the bank

ACADEMIC PROJECTS:

University of Connecticut, School of Business, Hartford, CT January 2016-August 2016

Auto Insurance Modeling, Travelers: Case competition to predict auto claim cost for each policy based on vehicle value, vehicle age, exposure to identify high risk customers thereby introducing rating plans. Built compound Poisson models, zero inflated compound Poisson models to accurately predict claim cost

Health Insurance and medical expenses prediction: By estimating risk of health care system, predicted annual medical expenditures based on predictors including age, annual income, gender, health, no of chronic diseases, physical limitation which gives a heads up on the estimated medical expenses and health insurance plans. Performed data modelling such as lasso and ridge regression to penalize parameters and determine factors that affect medical expenditures

Natural Language Processing: Analyzed viewer sentiments of movie reviews in order to gauge customer satisfaction. Recommend to producers what customers expect and how to create movies to maximize business. Performed text analytics and built decision tree to predict sentiments and understand significant words determining sentiment

Contact this candidate