Spearheaded a team of 9 data scientists and engineers with which have conceived a Multi Touch Attribution product from scratch launched as a cloud agnostic licensed product.Using AI and Data Science technology in the corporate world to overcome complex business challenges. A dedicated effort to contribute with excitement and differentiated innovation to our clients in order to demonstrate value and benefits. Among the contributions are advances in Predictive Analytics, Machine Learning, Deep Learning, Natural Language Processing and Computer Vision.Areas around Data and AI: - KPI driven approach- Machine Learning E2E pipeline using "collect" - "transform" - "analyze" - "monitor" strategy- ML explainability using LIME, SHAP- Extensive usage of GitHub,R, Python, CI/CD pipelines in MLOps for faster path to production using Docker, Kubernetes, Microservices-Usage of supervised learning techniques like Random forest/Ensemble Logistic Regression/Ridge and lasso regression/Time series forecasting/SVM/XGboost/Adabost and unsupervised algorithms like RNN/LSTM to solve pratical business problems.

Managing team of analytics consultants, analysts to deliver business strategy, Risk & Fraud operations and analytics support, Credit Risk strategy, Payments & transactional risk analytics to facilitate data-based decision-making• Support Risk Management & Data Analytics team with decision engines and BI platforms• Dashboards for Credit Risk and Analytics teams providing real time data access to support senior government executives in decision making• Conceptualizing and building the global risk and fraud analytics (descriptive and predictive) ecosystem• Deriving actionable trends and insights on global fraud landscape through analysis using big data tools

Key activities include:Segmentation & Behavioural AnalysisProgram Enhancement Recommendations (Prize Cost Optimisation,Reward milestones, etc.)Customer Lifetime Value AnalysisProgram/Campaign Forecasting and Impact AnalysisProgram KPI monitoringBest practice Recommendations based on Lumata experience and external benchmarkingAttrition management of existing customer by use of various predictive modelling techniquesExperience: SAS Base and Macro Language, SQL, Excel, VBA

Development of statistical model and scorecard using multi-bureau data to assess risk for credit card applicants with existing Current Accounts with American Banking clientCustomer churn modeling for American banking client (Defined attrition based on "inactivity" of customers, Text Mining, Sound and Text Matching, Triggers for Attrition)Developed logistic regression based predictive models and validated them using historical data; maintained the models and upgraded them continuously based on latest information and featuresDeveloped organizational resources for Non-Linear Regression & Classification techniques - Neural Networks, Random Forests, SVM & K-Nearest NeighborsLeveraged unsupervised learning techniques to cluster cards' and other transaction level data systematicallyDevelopment of algorithms & resources for Non-Linear Regression, Classification & Time Series ModelingDeveloped highly competent solutions using SAS, Tableau, VBA and SQL