Nvidia AI Architect- Delivery

Key Responsibilities :

• Lead end-to-end transitions of AI PoCs into production environments, managing the entire process from testing to final deployment.
• Configure, install, and validate AI systems using key platforms, including VMware ESXi and vSphere for server virtualization, Linux (Ubuntu / RHEL) and Windows Server for operating system integration,
• Docker and Kubernetes for containerization and orchestration of AI workloads.
• Conduct comprehensive performance benchmarking and AI inferencing tests to validate system performance in production.
• Optimize deployed AI models for accuracy, performance, and scalability to ensure they meet production-level requirements and customer expectations.
• Serve as the primary technical lead / SME for the AI POC deployment in enterprise environments, focusing on AI solutions powered by Nvidia GPUs.
• Work hands-on with Nvidia AI Enterprise and GPU-accelerated workloads, ensuring efficient deployment and model performance using frameworks such as PyTorch and TensorFlow.
• Lead technical optimizations aimed at resource efficiency, ensuring that models are deployed effectively within the customer’s infrastructure.
• Ensure the readiness of customer environments to handle, maintain, and scale AI solutions post-deployment.
• take ownership of AI project deployments, overseeing all phases from planning to final deployment, ensuring that timelines and deliverables are met.
• Collaborate with stakeholders, including cross-functional teams (e.g., Lenovo AI Application, solution architects), customers, and internal resources to coordinate deployments and deliver results on schedule.
• Implement risk management strategies and develop contingency plans to mitigate potential issues such as hardware failures, network bottlenecks, and software incompatibilities.
• Maintain ongoing, transparent communication with all relevant stakeholders, providing updates on project status and addressing any issues or changes in scope.

Experience :

Technical Expertise :