At Coupang, I lead infrastructure and AI/ML platform engineering for company-wide search ranking, recommendation, ads, and LLM workloads. My organization operates hybrid datacenter and AWS compute across 10,000+ GPUs, 50+ AWS accounts, and SLO-governed ML platforms.
I built GPU training, storage, and inference platforms spanning InfiniBand fabrics, Kubernetes scheduling, WEKA, Lustre, VAST Data, Triton, and vLLM. These platforms improved GPU utilization from 42% to 71%, reduced time-to-train by 60%, and serve 15+ production models at 80K QPS peak.
Previously at Intuit, I led AWS platform standards, multi-account governance, financial-platform migration, and resilience architecture across 200+ accounts. I helped deliver $25M in annual run-rate savings while improving deployment velocity and tax-season reliability.
I also led enterprise cloud security and DevSecOps engineering at Fidelity Investments, building policy-as-code, continuous compliance, and CI/CD reliability automation. Across AI infrastructure, AWS economics, and SRE, I partner with ML, finance, engineering, and executive teams on capacity, cost, resilience, and long-term platform roadmaps.
