India
Developed credit scorecard models using logistic regression to estimate customers' probability of repayment. Incorporated key behavioral and demographic variables, and applied feature selection, binning, and model validation techniques to ensure regulatory compliance, interpretability, and strong predictive performance
Improved website and vehicle experience for vehicle owners by developing a Natural Language Processing (NLP) Multi-class classification model, in collaboration with Ford US team, with an overall accuracy of 80% to provide actionable insights to the company. Developed Multi-class classification model with 80% accuracy to capture public perception and identify the key talking points of newly launched Ford cars in several markets.
Automated the New Car Buyer Survey reports for Ford Marketing team, UK,with the help of Alteryx workflows and Qlikview Dashboard reducing the turn-around time from 10 weeks to 2 weeks. The report helped to analyse car buyer’s feedback and identify areas of concerns and satisfaction. Lead the development of Marginal Gains approach for the employees PULSE survey thereby improving team’s PULSE scores from 70% to 90%. Influenced marketing campaigns aimed at improving brand image by capturing insights on Brand sentiments and Brand keywords for business teams across the globe using online administered surveys. Involved in market study for vehicle design and planning. Learnt how research insights are utilised in vehicle design and development. Developed ComScore Tableau Dashboard Demand-sensing and Cross-shopping that analyzed new automotive shoppers. It increased turn-around time and reduced costs for the company.