Karan Rajendra

Research Assistant – Distributed Systems & ML Infrastructure at Stony Brook University

Engineered a distributed ML data pipeline using Python and Kubernetes-based scheduling, accelerating model training throughput by 500 samples per second for the core research team. Architected fault-tolerant microservices with Kafka for event streaming and gRPC for communication, containerized in Docker to maintain data integrity during weekly chaos engineering tests. Optimized a large-scale distributed training job by profiling with perf and Valgrind, reducing its completion time from 5 hours to 4 hours without altering the model's accuracy. Automated a multi-node CI/CD pipeline with GitHub Actions and Prometheus monitoring, slashing environment setup time for new team members from 2 hours to 30 minutes.

Graduate Teaching Assistant at Stony Brook University

Instructed a cohort of 60+ graduate students on core distributed systems principles, including Raft, data sharding, and replication strategies for fault-tolerant architectures. Architected and debugged 5+ hands-on programming assignments using gRPC, requiring students to implement and analyze trade-offs in consistency, availability, and partition tolerance. Mentored students through complex system design challenges, elevating the cohort's average score from B- to B+ across 5 programming assignments and fostering collaborative skills.

Software Engineer at Dell Technologies

Spearheaded the design and implementation of Java-based microservices architecture using Spring Boot, enhancing system reliability and reducing bottlenecks for Dell’s multi-region enterprise infrastructure platforms. Revamped React.js front-end interfaces for cloud dashboards, improving user navigation and accessibility for thousands of internal and client users. Orchestrated integration of RESTful APIs with Swagger-documented endpoints, enabling seamless service-to-service communication and real-time insights for infrastructure monitoring and telemetry systems. Modernized PostgreSQL database queries and schema structures, reducing average latency by over 50% and enabling near real-time analytics for cloud resource utilization and reporting. Automated deployment of containerized applications using Docker and Jenkins CI/CD pipelines, ensuring zero-downtime releases, consistent build environments, and faster delivery cycles for internal software tools. Collaborated with cross-functiona

Research Assistant – Distributed Systems & ML Infrastructure at Stony Brook University

Engineered a distributed ML data pipeline using Python and Kubernetes-based scheduling, accelerating model training throughput by 500 samples per second for the core research team. Architected fault-tolerant microservices with Kafka for event streaming and gRPC for communication, containerized in Docker to maintain data integrity during weekly chaos engineering tests. Optimized a large-scale distributed training job by profiling with perf and Valgrind, reducing its completion time from 5 hours to 4 hours without altering the model's accuracy. Automated a multi-node CI/CD pipeline with GitHub Actions and Prometheus monitoring, slashing environment setup time for new team members from 2 hours to 30 minutes.

Graduate Teaching Assistant at Stony Brook University

Instructed a cohort of 60+ graduate students on core distributed systems principles, including consensus algorithms (Raft), data sharding, and replication strategies for fault-tolerant architectures. Architected and debugged 5+ hands-on programming assignments using gRPC, requiring students to implement and analyze trade-offs in consistency, availability, and partition tolerance. Mentored students through complex system design challenges, elevating the cohort's average score from B- to B+ across 5 programming assignments and fostering collaborative skills.

Software Engineer at Dell Technologies

Spearheaded the design and implementation of Java-based microservices architecture using Spring Boot, enhancing system reliability and reducing bottlenecks for Dell's multi-region enterprise infrastructure platforms. Revamped React.js front-end interfaces for cloud dashboards, improving user navigation and accessibility for thousands of internal and client users, while ensuring responsive, cross-browser compatibility. Orchestrated integration of RESTful APIs with Swagger-documented endpoints, enabling seamless service-to-service communication and real-time insights for infrastructure monitoring and telemetry systems. Modernized PostgreSQL database queries and schema structures, reducing average latency by over 50% and enabling near real-time analytics for cloud resource utilization and reporting. Automated deployment of containerized applications using Docker and Jenkins CI/CD pipelines, ensuring zero-downtime releases, consistent build environments, and faster delivery cycles. Collabo

Research Assistant – Distributed Systems & ML Infrastructure at Stony Brook University

Led design of distributed ML data pipeline in Python with Kubernetes scheduling, boosting core model training throughput by 500 samples/second. Built fault-tolerant microservices using Kafka for event streaming and gRPC for inter-service communication, containerized with Docker to preserve data integrity during chaos tests. Optimized a large-scale distributed training job with perf and Valgrind, cutting total runtime from 5 hours to 4 hours without impacting accuracy. Implemented a multi-node CI/CD pipeline with GitHub Actions and Prometheus monitoring, reducing environment setup time for new team members from 2 hours to 30 minutes.

Graduate Teaching Assistant at Stony Brook University

Instructed 60+ graduate students on core distributed systems principles, including Raft, data sharding, and replication strategies for fault-tolerant architectures. Architected and debugged 5+ gRPC-based hands-on assignments, requiring students to analyze trade-offs in consistency, availability, and partition tolerance. Mentored students through complex system design challenges, elevating cohort average from B- to B+ across 5 programming assignments and fostering collaborative skills.