● Combined RFM segmentation with survival analysis to predict Customer Lifetime Value score resulting in over 15% increase in customer retention rates.
● Created an Isolation Forest-based anomaly detection system to identify potentially fraudulent transactions, enhancing fraud detection accuracy by 10% and minimizing financial risks.
● Built a simple exponential smoothing forecasting model to accurately predict future revenues, enabling the business heads to make informed resource allocation and budgeting decisions.
● Utilized SQL, Pandas and NumPy to extract, manipulate and clean the data by removing the outliers, duplications, and inconsistencies.
● Developed Tableau dashboards to track KPIs and data-driven decision-making, automating 15 hours of reporting work weekly.
● Designed ETL pipelines using SSIS for HR, accounting, and transaction data, resulting in streamlined data movement and improved system performance by over 70%.
● Developed a robust data warehouse in SQL Server, leveraging data from CSV, Excel sheets, and Oracle Server, ensuring clean and accurate data for analytical models.
● Utilized SQL, Pandas and NumPy to extract, manipulate and clean the data by removing the outliers, duplications, and inconsistencies.
● Constructed integrations in Python to automate scraping, cleaning, and migration of IFSC codes, saving over 5 hours of manual work every week.
● Employed a Django-based monitoring solution to track key metrics of production servers (CPU Utilization, Memory, Hits/sec, Response time) in real time.