Providing technical oversight of distributed engineering teams in India and the USA for a unified risk platform consolidating fraud detection across gift card channels. Re-architected XGBoost-based ML services integrated into a centralized risk engine. Designed a real-time data aggregation framework using DynamoDB and Lambda architecture, and developed AWS Batch-based data injection pipelines. Delivered detailed design of a high-performance, Drools-based rule engine. Exploring Generative AI tools like Claude to enhance fraud decisioning.

Architected and developed a real-time Flink streaming application on AWS EKS to ingest and process data from NFS to Snowflake. Built and managed Airflow DAGs to orchestrate batch workflows within the data lake. Designed and implemented data privacy controls to remove opt-out accounts from multiple data sources. Collaborated closely with data scientists to develop scalable machine learning pipelines using AWS SageMaker.

Technical Team Leader and Architect with deep expertise in designing real-time, high-scale systems and deploying machine learning in production. Architected and led the development of a low-latency fraud detection platform processing 30M+ daily transactions using Spring Boot, distributed Linux, MQ, ELK, Cassandra, and XGBoost—earning a global innovation award.

Partnered with data scientists to source training data across banking channels and built a Java framework to convert XGBoost PMML models into efficient native code.

Designed a multi-cluster Cassandra architecture for real-time and batch ML pipelines, and drove cloud-based ML POCs using AWS services (S3, Lambda, SageMaker).

Recognized for leading cross-functional teams, setting technical direction, and delivering innovative, production-grade solutions at scale.

Led the detailed design and development of various risk and fraud technology projects, including real-time fraud detection middleware processing credit card transactions from core banking systems using Spring Integration (SEDA), Drools, and distributed caching; predictive analytics integration with big data platforms across credit and debit channels leveraging distributed computing and Cassandra; and a scalable travel prediction scoring engine supporting real-time decisioning. Also mentored and led a high-performing technical team through end-to-end solution delivery.

Detail design and development of auto loan workflows on the Capital One 360 platform, integrating RTM middleware services. Implemented KYC, Early Warning Systems (EWS), and advanced security authentication mechanisms for the retail banking web platform.