Venkata Upendar Reddy Kamasani

• Partnered with risk analysts to enhance credit risk scoring models, improving decision consistency and reducing manual review rates by ~10%.
• Contributed to real-time fraud detection workflows using Kafka + Flink, reducing detection lag from minutes to seconds.
• Automated ML release processes with Jenkins/GitLab CI, cutting deployment steps from hours to under 30 minutes.
• Containerized ML workflows with Docker + Kubernetes, improving scalability of production systems.
• Supported customer segmentation models that informed targeted marketing campaigns.

• Applied ARIMA and Prophet forecasting models, improving short-term inventory accuracy by ~12%.
• Built analytics pipelines with Spark + Kafka to monitor supply chain events in near real time.
• Deployed lightweight ML microservices that helped pharmacovigilance teams monitor adverse events.
• Consolidated R&D, clinical, and commercial datasets into Snowflake, reducing reporting turnaround time from days to hours.
• Supported statistical validation for FDA submission datasets, ensuring regulatory compliance.