Hafiz Muhammed

Product Profitability Framework - Personal Banking

• Developed a ‘Target Prioritization Framework’ in order to reduce the problem of ‘Empty Widgets’ (unutilized product) and to increase the utilization of Unsecured Line of Credits. Leveraged MARS to build propensity models for acquisition, utilization and Random Forest to predict balances to identify customers who will generate a potential $20M in profitability in 3 years

Digital Primacy - Personal Banking

• Adopted a rule based approach to define levels of ‘Digital Engagement’ of personal banking customers in order to identify potential customer-product combinations which will entrench digital customer base leading to an yearly incremental profit of $1.6M
• Used K-Means Clustering to segment customer universe based on their digital behavior and fitted a decision tree on segments to identify rule sets to define customer’s digital primacy

Onboarding Customer Profiling - Search Engine Marketing

• Developed a segmentation model to categorize newly onboarded customers into segments in order to understand their need state and helping the marketing team to develop customized pitches leading to lower churn and higher monetization during onboarding phase
• Used K-Means Clustering to segment the Onboarded customers based on their performance during initial 15 days and created a production pipeline to implement this in real time

Customer Vertical Classification - Search Engine Marketing

• Developed an automated scalable classification model to predict the line of business of a customer based on it’s performance patterns in order to avoid manual processing
• Used Random Forest to create a model to predict the customer’s business vertical