• IArchitected a scalable Go/Python microservices stack via gRPC powering global recommendation infra. • Implemented a multi-stage recommendation engine with Two-Tower, DeepFM & LightGCN models on PyTorch. • Developed Actor-Critic RL agent enabling dynamic, multi-objective ranking and throughput optimization. • Provisioned end-to-end cloud infrastructure on AWS via Terraform (EKS, S3, MSK Kafka, IAM policies VPCs). • Containerized 10 microservices in Docker; orchestrated deployment and autoscaling via Kubernetes and Helm. • Engineered real-time Feature Store on Kafka & Redis Cluster, and an Apache Iceberg data lake on S3.