NIM Solution Architect
Nvidia · Shanghai
NVIDIA is leading company of AI computing. At NVIDIA, our employees are passionate about AI, HPC , VISUAL, GAMING. SA team is more focusing to bring NVIDIA new technology into difference industries. This role focuses on NVIDIA Inference Microservices (NIM), inference / RL rolloutperformance, and AI workflow enablement for LLM, VLM, and other generative AI workloads. It is a highly hands-on position at the intersection of model optimization, inference infrastructure, and customer solution delivery.
What you’ll be doing:
- Drive the implementation, deployment, and optimization of NVIDIA Inference Microservices (NIM) solutions for enterprise and industry AI workloads.
- Package and serve open-source, NVIDIA, and customer-proprietary models through NIM with standardized, containerized APIs for on-premises, cloud, and hybrid environments.
- Optimize high-volume inference and rollout workloads for LLMs and VLMs.
- Evaluate and tune the NIM models.
- Deliver technical projects, demos and client support tasks as directed by the Solution Architecture Leadership.
- Provide technical support and guidance to customers, facilitating the adoption and implementation of NVIDIA technologies and products.
- Collaborate with cross-functional teams to enhance and expand our AI solutions portfolio.
What we need to see:
- Master’s degree or higher in Computer Science, Machine Learning, Electrical Engineering, Mathematics, or a related technical field, or equivalent experience.
- 2+ years of hands-on experience in machine learning engineering, applied research, LLM/VLM inference, or RL rollout.
- Production-quality Python and PyTorch skills, including distributed GPU training, solution, profiling, debugging, memory optimization.
- Working knowledge of transformer architectures, performance optimization, rollout sampling strategies, structured generation, and model-quality evaluation.
- Strong written and verbal communication skills, with the ability to collaborate effectively across research, engineering, infrastructure, product, and customer-facing teams.
Ways to stand out from the crowd:
- Publications, open-source contributions, or significant technical projects, LLM/VLM, agent systems.
- Experience applying programmatic verification, simulators, compilers, execution sandboxes, APIs, or external tools as reward sources for model training. agent system.
- Familiar with oss RL framework such as SLIME, Nemo-RL.
- Familiarity with enterprise AI deployment, customer adaptation, or adapting foundation models to specialized vertical domains.