Feature Engineering, Retention Strategy, Prediction with 85?curacy, Customer Life Time Value, A/B testing

Authorization, Fraud and Chargeback diagnosis of a major UK bank vis-à-vis Market.

Created algorithms (BagofWords and Naïve Bayes classifier) for analyzing consumer’s opinions about product and brand.

Evaluate impact of advertising investment on retail business in the US.

Built a product for direct sales industry (180 billion dollar) to improve customer engagement, reduce churn and increase revenue with minimuminvestments.Involved in providing Thought Leadership, Product Implementation, Data Analytics and Client Management.Key contributor in fetching seed funding for DirecTech Labs.

Feature Engineering & Implemented Ensemble of machine learning algorithms to predict the probability of churn of a customer.Targeted top 20% high risk customers with the right retention techniques.

Designed engagement platform utilizing the power of Chat Bots, Email, SMS and Web Apps.Based on Customer Lifetime Value, right channel was assigned to target and engage these customers.

Built influence score of each customer. (Similar to credit score in Banking industry)Based on the influence score (Derived from key Factors) of the reps, task was assigned to them to save at risk reps in their network.

Provided clients with a new perspective on their customer base.This enables clients to truly understand who is building their business and who is negatively impacting their business.

Different features were introduced in the product and launched for the entire population after successful A/B testing.Results of A/B testing has been measured using Frequentist and Bayesian approach.

Objective: Credit Card Portfolio Analysis Behavioral Segmentation Profitability AnalysisActionable Results: Analysis helps in designing segment specific yearly marketing plan and credit line management strategies.

Objective: Analysis of major UK bank performance with emphasis on transaction declines.Actionable Results: Areas of opportunities were identified for client in terms of low performing Merchant Category.Areas of Major Fraud along with Ticket size are identified for better Targeting.

Objective: Analyzing positive and negative chatter on Twitter and Face bookResponsibility:Created algorithms (Bag of Words and Naïve Bayes classifier) for analyzing consumers opinion about product and brand.Extensive research and development of algorithm for Arabic sentiment analysis.

Objective: Churn Management and Retention Strategizing Non-Payment Prevention Responsibility:Conceptualized and developed a predictive model to identify prospective churners for prepaid and postpaid business.Identify customers likely to have non-payments in next 90 days.Collaborated with marketing team to come up with right strategies and offers for retention of customers.

Objective: Evaluate impact of advertising investment on retail business in the US.Responsibility:Lead a team of 6 members to deliver unobserved component and market mix models.Compute ROI of media activities and identified marketing drivers that had greatest impact on sales.Lead a team of 4 members and develop their skill set in designing data warehouse that support modeling.

Responsibility:Employed nonlinear programming algorithms to optimize incremental sales of marketing drivers.Created business monitoring dashboard and tools for optimization through VBA and SAS.Simulation of changes in ROI through what-if scenarios such as shifting spending allocation among marketing activities.

Objective: Measure impact of media investments on brand awareness and Compute Long Term ROI of Marketing Drivers.Responsibility:Compute Brand health score from the survey data.Employed clustering (PROC VARCLUS) to identify brand health metrics which drives brand sales.

Objective: Design effective pricing strategy and improve firms profitability.Responsibility:Provided insights on potential pricing and promotion strategy of the different products.Checked the Effectiveness of the in-store promotions and used Log-linear regression to get price elasticity.

Objective: Measure Performance of Investigator in Clinical TrialsResponsibility:Conceptualized Regression model on Performance of Investigator in Clinical Trials.Used SAS, SAS Enterprise Miner and Knowledge seeker to attain essence of data across various databases