I am an aspiring AI infrastructure and backend engineer with a Computer Science Honours BSc from the University of Toronto and upcoming Master of Engineering studies in Electrical and Computer Engineering at the University of Waterloo.
I build backend services and stateful AI agent workflows using technologies such as OpenAI APIs, gRPC, GraphQL, SQLc, Python, Go, and SQL. My work focuses on reliable agent orchestration, RAG systems, evaluation pipelines, and production-oriented API development.
I have hands-on distributed systems experience from internships and research, including cloud scheduler development, Kubernetes-based test automation, RDMA networking benchmarks, and Spark Shuffle transport integration. I am comfortable working with Linux, Docker, Kubernetes, TCP/UDP networking, and high-performance data-transfer systems.
I have also applied machine learning to infrastructure problems, training LSTM and KNN models on workload traces for dynamic resource scheduling. My project work includes multi-agent data exchange systems, ETL verification workflows, and speculative decoding research for large language models.
I value measurable engineering improvements and use evaluation metrics such as precision, recall, throughput, latency, task success rate, test coverage, and cost reduction to guide iteration. I hold the Certified Kubernetes Administrator certification and am interested in AI infrastructure, distributed systems, backend engineering, and LLM serving.
Developed stateful agent workflows and backend APIs using the OpenAI API, gRPC, GraphQL, and SQLc.
Built a Prometheus-based evaluation pipeline to measure precision, recall, and task success rates, driving prompt and workflow iteration.
Built RDMA-enabled network interface benchmarks to characterize communication performance and compare it with socket-based TCP and UDP baselines.
Integrated UCX/RDMA transport into a Spark Shuffle plugin, deployed multi-node experiments, and measured throughput and latency for large-scale data transfer.
Developed and refactored a Huawei Cloud scheduling engine, adding trait-based resource-provisioning strategies for heterogeneous cluster workloads.
Built Kubernetes and Docker test environments and automated scheduler testing, increasing test coverage to 80%. Trained LSTM and KNN models on Azure workload traces for dynamic resource scheduling, reducing user cost by 20% at comparable performance.
Master of Engineering, Electrical and Computer Engineering
Incoming Master of Engineering student in Electrical and Computer Engineering.
Bachelor of Science, Honours, Computer Science
CGPA: 3.63/4.00. Received Dean's List, Archibald Macmurchy Memorial Scholarship, and University College Special Curriculum Scholarship.
This professional hasn’t added portfolio projects yet.
This professional hasn’t listed any services yet.
