Architect end-to-end GPU cluster design for a next-generation AI cloud platform, making decisions across compute, networking, and storage that power LLM training and inference at the scale of tens of thousands of GPUs. Core work spans InfiniBand/RoCE interconnect topology, GPU architecture (NVIDIA/AMD), performance modeling, and automation tooling.
- Build performance benchmarks and tuners for NCCL and PyTorch - Develop AI infrastructure and training components for large scale models - Enable reliable scalable distributed ML training - Improve distributed GPU…
Senior AI/ML engineer building RAG pipelines, LLM-based applications, semantic search, and full-stack features including backend services, REST APIs, and front-end UIs, with MLOps, prompt engineering, and Azure DevOps CI/CD.
Senior engineer building AI-driven cloud compliance automation—crafting Rego policies, Terraform modules, and RAG-powered agentic workflows that integrate LLM APIs to detect configuration drift.
Platform engineer at ai&, building the serving layer between bare metal and inference engines across GPU clusters (AMD, NVIDIA, Tenstorrent) in company-owned data centers. Owns Kubernetes orchestration, scaling, RDMA networking, observability, and CI/CD on a heterogeneous compute fleet.