
Python Data Scientist
Hexaware Technologies · Feb 2023 – Present
- Contributed to development of an enterprise Al platform processing 500,000+ daily transactional records using PySpark and Spark SQL.
- Improved Customer Churn Prediction model accuracy to 88% through advanced feature engineering and hyperparameter tuning.
- Recognized with the Outstanding Contributor Award for enhancing model performance and optimizing inference efficiency, reducing cloud compute overhead.
- Developed personalized recommendation models to support cross-sell and revenue optimization strategies.
- Designed and deployed real-time ML inference APIs using FastAPI and Docker for seamless integration with business dashboards and admin systems.
- Built churn prediction models achieving 84% accuracy to support proactive customer retention initiatives.
- Developed NLP pipelines to analyze customer complaints and support tickets for sentiment analysis and issue classification.
- Engineered scalable ETL pipelines for processing telecom datasets including CRM, billing, and network logs.
- Supported Al-driven routing optimization models for improving network performance and reducing service disruptions.
- Designed fraud detection and anomaly detection models for real-time transaction risk scoring.
- Processed large-scale transactional datasets using PySpark for fraud pattern identification and risk profiling.
- Built RESTful APIs using Django REST Framework to serve fraud risk scores to operational systems.
- Contributed to reducing manual verification effort by deploying automated classification models.

