Arunkumar Veeramalli

Business Data Analyst @Citi

Boca Raton, FL, US

Developed a data pipeline using AWS Glue and Python (pandas, NumPy) to ingest and clean 30 million customer records from the Virtual Enterprise Data Lake, reducing data preparation time and supporting timely churn model development. Wrote complex SQL queries on Snowflake and SQL Server to consolidate multi-source data, including transaction logs, CRM feeds, and support interactions, creating consistent input datasets for modeling and segmentation. Built a K-Means clustering model in Python to segment customers by behavior and tenure, identifying five actionable groups that improved targeting and boosted retention campaign response rates by 38%. Created a logistic regression churn model using Python (scikit-learn), achieving 92% accuracy. Collaborated with model governance teams to ensure full compliance with Citigroup\'s MRM and AIM frameworks. Automated data validation with Python and SQL to monitor source reliability and schema changes, cutting manual QA time by 60% and improving model trustworthiness. Designed and deployed Tableau dashboards to track churn KPIs, campaign ROI, and customer lifetime value by region and product. Dashboards were used in monthly strategy meetings and influenced over $10M in marketing spend. Led A/B testing across four retention campaigns, defining control/treatment groups based on churn risk. The insights drove a 22% increase in campaign effectiveness and optimized contact frequency. Investigated key attrition drivers using RCA techniques in Python and SQL, revealing fee sensitivity and reduced credit utilization as primary factors. Findings led to revised incentives and a 12% quarterly churn reduction in the credit card segment. Coordinated with cross-functional teams in marketing, CRM, and compliance to align analytics with business objectives. Documented methodologies and model assumptions for internal audits and executive reporting.