Data Analyst

About the Role

A Data Analyst position with Franklin Templeton, requiring 0-4 years of experience, with an undisclosed salary.

What You Should Know

• •Credit scorecards (logistic regression and similar models) as the classic banking ML use case
• •Fraud detection - why precision/recall trade-offs matter more than raw accuracy
• •Model governance and explainability, since RBI increasingly expects explainable credit decisions
• •Dimensional modeling and slowly changing dimensions for customer/account history in a banking warehouse
• •MLOps concerns specific to regulated environments - model versioning, audit trails, drift monitoring

Interview Questions

1. 1.How would you evaluate a fraud detection model where false positives block genuine customers?
2. 2.What's the difference between accuracy and recall, and why does it matter more in credit risk models?
3. 3.How would you design a churn-prediction model for a bank's savings account customers?
4. 4.What data governance concerns are unique to working with banking/PII data?
5. 5.What GenAI use case would you propose for a retail bank, and how would you measure its success?

Scenario-Based Questions

1. 1.Your fraud model's false positive rate has been climbing, frustrating genuine customers. How do you diagnose and fix it?
2. 2.Leadership wants a GenAI chatbot live in 6 weeks. What would you scope in vs cut for v1?
3. 3.A model that performed well in testing is underperforming in production. What's your debugging checklist?

Interview Tips

• •Ground every model discussion in a business metric (bad debt reduced, fraud loss avoided, TAT improved) - banking leadership cares about rupee impact, not just AUC.
• •Be ready to discuss model risk management even for junior analytics roles - it's a real regulatory concern in banking.
• •If GenAI is in the JD, have a point of view on hallucination risk in a regulated, customer-facing context.