Vijayashree Marupeddi

Experimentation for personalized membership offers at Petco - Partnered with cross-functional teams to define key performance indicators, design membership offers, track and measure campaign performance.

• Collaborated on a Diabetic Comorbidity Prediction Research Project for a leading hospital chain in Cincinnati. • Developed a predictive model utilizing ensemble tree methods: Random Forest and XGBoost in scikit-learn. Conducted feature engineering on HIPAA data to create clinical biomarkers and assess social determinants of health.

Lubricants Sales Forecasting & Monitoring - • Developed a forecasting model in Python to predict the next purchase date of lubricants for tanks at customer locations across US by utilizing historical purchase order data • Built a comprehensive eCommerce decision board in Power BI capturing ecommerce lubricant sales, revenue along with marketing attribution for North - Americas Region to provide a holistic picture of eCommerce performance Counterparty Deduping • Developed a pyDash (Python) tool to deduplicate counterparties of various business systems through record linkage algorithm and Jaro Winkler technique. Enabled effective attribution of financial figures *Roles & Responsibilities*: • Part of the talent recruitment team - Interviewed candidates for analyst and associate data scientist roles • Developed and streamlined onboarding of new joiners into the team • Participated in leadership initiatives to develop reproducible analytical tools and algorithms

Supply Chain Optimization & Working Capital Reduction - • Built a linear optimization model in Python using CPLEX solver for freight planning & optimization across ports in Singapore • Delivered a user driven automated model detailing optimal stock levels to be maintained at various stock out % scenarios/service levels and identified opportunities for working capital reduction. Enabled $2M USD cash release • Collaborated with Directors & Managers and delivered an exhaustive Spotfire dashboard to monitor the performance of supply operations. Enabled better planning, fewer stockouts and cash benefits of $4.2M USD

Client: Fortune 100 Airline Carrier Social Media Analysis - Customer Care Effectiveness & Sentiment Analysis: • Developed a sentiment classifier (60% accurate) in Python for social media customer care team to understand customer sentiments on services / policies & address their grievances better • Identified common issues / queries of loyalty program's credit card customers by performing text mining in Python and delivered insights to the co-brand team to revise their offers Client: Fortune 300 Utilities & Services Provider Real Time Predictive Maintenance - Condenser Issues Detection: • Led a global team of 4 in developing a multi-class classifier in Python (XGBoost) using IoT sensor water telemetry data and condenser performance data • Predicted condenser issues across power plants to raise early alarms, resulting in optimal operation of condensers leading to reduced downtime, and reduction in GHG emissions Client: Fortune 100 US Telecom Provider - Customer Segmentation: • Profiled online customers based on clickstream data using K-prototype clustering technique to understand their behaviors and personalize web merchandising *Roles & Responsibilities*: • Mentored 10+ new inductees on problem solving methodologies and client communication • Trained ~10 Global Associates and 15+ delivery resources in R programming language • Involved in driving organizational initiatives across delivery teams as part of Leadership’s first team

Client: Fortune 100 Airline Carrier Campaign Analytics - Measurement & Automation • Performed Test - Control segmentation of target audience, carried out campaign measurements using different matching methods for accurate revenue attribution and performance evaluation • Built a propensity model in R (XGBoost) to predict the likelihood of customer conversion for marketing campaigns & improve audience targeting thereby enabling client devise better Root Cause Analysis - for IT issues causing flight delays • Identified top IT issues and their causes that accounted for 40% of departure delays across all US hubs in 2017 through exploratory analysis. • Analyzed resolution data using text mining to derive insights around the most frequent resolutions for different IT issues Route Specific Targeting (A/B Testing) • Analyzed performance of corporate companies across underperforming routes and identified target companies to roll out personalized offers • Implemented PCA on key business KPIs and clustered the corporate companies to determine accurate control companies Client: Fortune 100 US Telecom Provider Web Merchandising • Profiled online customers based on clickstream data using K-prototype clustering technique to understand their behaviors and personalize web merchandising thereby leading to higher conversion in lower cycle time

Master's with *Data Sciences Certification* Data Mining Algorithms, Machine Learning & Systems Design, Statistical Modeling, Probability Models, Forecasting Methods, Optimization, Forecasting Methods, Big Data Integration, Data Wrangling (R & Python), Simulation Modeling

Secured 48th University Rank. Graduated with distinction. COURSES - Object Oriented Programming, Artificial Intelligence, Database Management Systems, Data Structures, Probability & Queueing Theory, Java Lab

12th Grade: 93% 10th Grade: 9.4 GPA/10