Oracle is seeking an OCI GPU Black Belt to drive customer success in designing, deploying, and optimizing large-scale AI and HPC workloads on Oracle Cloud Infrastructure (OCI). This role combines deep technical expertise in NVIDIA GPUs, distributed training and inference frameworks, benchmarking and performance tuning, RLHF pipelines, and end-to-end solution delivery. The OCI GPU Black Belt will work in close collaboration with our sales, marketing, and technical teams to drive revenue growth and accelerate market penetration for our NVIDIA GPU compute services. The role will also include progressing business opportunities, delivering technical workshops and demonstrations, and supporting proof-of-concepts to drive cloud consumption and overall revenue growth.
What You’ll Do
- Engage directly with customers to understand their requirements for NVIDIA GPUs on Oracle Cloud to run AI infrastructure, graphics, and HPC workloads.
- Lead the solution design within a collaborative virtual team, mapping requirements onto Oracle cloud services and providing hands-on support during the proof-of-concept phase.
- Deliver technical workshops, proofs-of-concept (PoCs), and demos, collaborating closely with sales, engineering, and customer teams to validate end-to-end solutions and accelerate cloud adoption.
- Optimize end-to-end AI workloads by analyzing hardware bottlenecks and tuning parallel libraries for peak efficiency.
- Deploy and scale HPC clusters, configuring compute nodes and shared file systems to meet performance SLAs.
- Lead the architecture and deployment of scalable inference platforms, leveraging containerized microservices on Kubernetes and OCI GPU instances.
- Design and implement distributed training pipelines using frameworks like DeepSpeed and Fully Sharded Data Parallel (FSDP).
- Develop benchmarking and profiling solutions to measure training and inference performance.
- Guide customers in model selection and evaluation, optimizing cost and performance.
- Contribute to Oracle’s internal expert community, document best practices, and mentor peers on AI infrastructure design.
- Stay current with emerging AI infrastructure technologies and represent Oracle at industry events.
Requirements
- 5+ years of hands-on experience in AI/ML infrastructure or HPC, architecting and operating large-scale GPU-accelerated environments for training and inference.
- Deep proficiency with NVIDIA GPU technologies (CUDA, cuDNN), RDMA networking, and cluster orchestration tools.
- Expertise in distributed training and inference frameworks: PyTorch, TensorFlow, DeepSpeed, and FSDP.
- Strong background in performance optimization techniques.
- Familiarity with cloud-native practices: Docker, Kubernetes, Terraform, and CI/CD for infrastructure.
- Solid understanding of cloud architecture principles on OCI or comparable public clouds.
- Bachelor’s or Master’s degree in Computer Science, Electrical Engineering, or a related technical field.
This role offers the opportunity to shape Oracle’s AI/ML portfolio, drive revenue growth through technical leadership, and collaborate with customers to unlock the full potential of GPU-accelerated AI and HPC solutions.
