Feb 2022 – May 2022 · 3 mos · Erode, Tamil Nadu, India

• Relocation from the US to India, Wellbeing and Family Caregiving

Sep 2020 – Jan 2022 · 1 yr 4 mos

• Microsoft 365 Commercial, Data Science & Analytics
• Drive product usage by building end-user and customer centric propensity models through Machine Learning / Deep Learning algorithms in Azure Databricks (Spark environment)
• Provide guidance to Finance / Marketing teams on opportunity space and expectations by key geographies based on the propensity models for target setting / strategy.
• Study customer adoption journey and identify points of adoption acceleration, saturation and stagnation for interventions.
• Provide insights on change in working trends using product telemetry during the pandemic period (covid'19).
• Identify personas based on product usage behaviors to enrich insights about users/customers.
• Build next best actions models to recommend products based on likelihood to move up in the personas segments.
• Part of the team that won the "Mental Health Hackathon Challenge" in Sep'20.

Jan 2019 – Aug 2020 · 1 yr 7 mos

• Microsoft 365 Commercial, Data Science & Analytics
• Conducted multivariate retention analysis to define metrics that lay emphasis on high value scenarios of Office 365 product usage.
• Measure customers digital transformation through adoption of various products in O365 suite with emphasis on productive behaviors/actions.
• Cross-product recommendation engine; drive product adoption and site engagement. Built propensity models and collaborative filtering models to provide next best actions.
• Measure the impact of product adoption on cross-product usage and shift in user behavior through observational studies to infer causality.
• Deliver insights around product usage to drive adoption by comparing customers to similar clusters.
• Presented a poster on Causal Inference in Microsoft Machine Learning and Data Science conference.
• Key Solutions Delivered: Causal Impact Observational studies, Behavioral Analytics, Retention & Growth models, Product Recommendation Engine
• Techniques: Decision Tress, Mixed Linear Effects models, Logistic Regression, Propensity Models, Propensity Score Matching, Collaborative Filtering, Double ML, Xgboost
• Technologies: R, Cosmos/Scope, Azure Databricks, Python

Oct 2014 – Dec 2018 · 4 yrs 2 mos

• Business Performance, Planning & Campaign Insights for a leading Redmond based technology firm
• Responsible for Goal Setting & Budgeting for a loyalty program with roughly $1 billion Rev inflow
• Define & Monitor business metrics through scorecards & dashboards
• Providing detailed analysis on metric trends and drivers
• Forecasting KPIs, Opportunity & Risk sizing for program initiatives
• Understand User behavior, retention and churn based on acquisition channels
• Target users based on their behavior and likelihood to respond to marketing interventions
• Measure ROI for various marketing interventions (A/B Testing)
• Key Solutions Delivered: A/B Testing, Behavioural Analytics, Business Outlook & Forecasts, Opportunity & Risk Sizing, Audience Segmentation and Targeted Acquisitions
• Techniques: Statistical Hypothesis Testing, A/B Testing, Univariate / Bivariate Analysis, Time Series Forecasting, Business Models, Simulations.
• Technologies: SQL, COSMOS/SCOPE, C#, R, Power BI, Excel, PowerPoint
• Delivered predictive models to help improve patient care and services for a leading Healthcare Service Provider in the APAC region. Helped identify patients with high 30d re-admission risk leading to cost savings of around 8%.
• Solutions Delivered: Readmission Risk Prediction
• Techniques: Feature Engineering, Univariate / Bivariate Analysis, Logistic Regression and Decision Trees
• Technologies: R

Apr 2013 – Sep 2014 · 1 yr 5 mos

• Worked with the Central Marketing Team of a leading technology firm to improve the effectiveness and efficiency of marketing activities by providing actionable insights using survey and social media.
• Key Solutions Delivered:
• Audience Segmentation & Micro-Targeting - Identified groups of customers who are most favorable to own vs. competitor products. Also, identified segments of users who neither lean towards own & competitor products for targeting.
• Ad Spend Optimization - Optimized the marketing spends of around $500 million across media channels and geographies; resulted in re-allocating the marketing budgets across channels and reducing where required.
• Brand / Product Drivers Analysis - Understood key marketing messages and product imagery perceptions that help drive and boost the brand image.
• Touch-Point Analysis - Responsible for identifying marketing channels and sources of information that help users convert from the awareness to purchase stages of the user funnel.
• Other Ad-hoc Analysis: Social Conversations & Online Reviews Mining, Switchers & Returners Analysis using Twitter Data, Survey Standardization & Audits.
• Techniques: Statistical Hypothesis Testing, Feature Engineering, Univariate / Bivariate Analysis, Clustering, Linear & Logistic Regression with transformation, Decision Trees and Linear Optimization
• Technologies: SAS, R, SPSS, Excel, PowerPoint, Azure ML

Nov 2010 – Mar 2013 · 2 yrs 4 mos

• Provide predictive analytics solutions on internal and external web properties of a leading internet services provider; Comparison Shopping, e-Commerce giant
• Key Solutions Delivered:
• Online Deal Classifier - Improved the accuracy of deal classification through a automated classification model using text mining, regex and decision tree algorithms. 12% improvement in the overall classification accuracy leading to eliminate the cost incurred on manual classification by crowd-sourcing internet marketplace.
• AdWords Bid Optimization (SEM) - Responsible for building an automated bidding engine that bids for keywords on search engines learning on a near real-time basis from past data and keyword performance. Worked with the data engineering teams to integrate the process into the BAU process. Resulted in a increase in revenue of 10% - $10 million revenue and and profit of $2 million yearly.
• Search Listing Optimization - Optimized the search listings ranking within the eCommerce portal to maximize the CTR and revenue. Led to an increased CTR of 15% from the baseline and a revenue increase of 8%.
• Techniques: Feature Engineering, Univariate / Bivariate Analysis, Regex, Statistical Hypothesis Testing, Logistic Regression, Random Forests and GBM
• Technologies: ORACLE/SQL, SAS, R, Excel, PowerPoint, Tableau

May 2008 – Oct 2010 · 2 yrs 5 mos

• Delivered an optimization framework for enabling the lead generation teams to meet targets in Leading Financial Services Providers' online portal. Delivered an end-to-end solution for Ad-Inventory Optimization resulting in a lift in leads of 18% (~2k leads a month).
• Techniques: Linear Optimization & Regression.
• Forecast Sales for early Supply / Demand Planning for Leading Consumer Electronic & Technology Giant
• Solutions Delivered: Forecast Market Volume / Predict Market Share, Forecast Retail Chain Sales by Department
• Techniques: Time Series models using an ensemble of methods Singular Value Decomposition, Linear models, ARIMA and Exponential Smoothing.
• Modeled customers likelihood of delinquency for a Peer-to-Peer lending portal based on the customer demographics; credit score; recent payment behavior and user profile information on the portal
• Technologies: R/Excel Interface, SAS, PowerPoint, Google Web ToolKit

Jul 2007 – Dec 2007 · 5 mos · Hosur · On-site

• Export Warehouse Layout Optimization and Inventory Management