About the job
NVIDIA is seeking a highly experienced and passionate Senior Software Engineer to join a team building large-scale machine learning solutions, including Graph Neural Networks, Tabular Foundation Models, and ensemble models. This role is critical to accelerating PyTorch-based frameworks and supporting user-facing tools that power NVIDIA’s cutting-edge data science solutions. The team works at the intersection of high-performance computing, GPU acceleration, machine learning infrastructure, and customer-facing software. This is an opportunity to shape efficient training and inference workflows for advanced machine learning models running on NVIDIA GPU infrastructure. If you are passionate about building high-performance software and enabling customers to solve complex data science problems at scale, we would love to hear from you. NVIDIA teams work on some of the world’s most ambitious computing problems, helping organizations adopt AI technologies that enable faster, smarter decisions.
Responsibilities
Develop accelerated, PyTorch-based solutions for large-scale machine learning models, including GNNs, TFMs, and ensemble models, with a focus on efficient training and inference on GPU infrastructure
Support CUDA-X Libraries and integrations used in PyTorch-based, large-scale machine learning workflows
Partner with developers, product managers, and scientists to develop innovative GNN models and GPU-accelerated implementations for model development and prediction phases
Develop solutions that help customers adopt NVIDIA hardware and software, and gather technical requirements directly from customers and Solutions Architects to guide product and engineering priorities
Provide technical leadership and mentorship to engineers across the team
Identify opportunities to improve the codebase and reduce code-maintenance overhead through re-architecture
Apply agentic coding tools to identify and fix bugs, implement new features, and refactor code
Solve complex technical issues, explain solutions clearly, exercise technical leadership, and coordinate across multiple teams to achieve shared objectives
Qualifications
Minimum
Bachelor’s degree (or equivalent experience) plus 5 or more years of relevant experience in large-scale machine learning, deep learning, and general data science; or a Master’s degree or PhD plus 3 or more years of relevant experience
3 or more years of experience with PyTorch
2 or more years of experience training enterprise-scale machine learning models across distributed infrastructure
2 or more years of experience designing and operating efficient training and inference workflows on GPU infrastructure, including profiling, scaling, orchestration, and resource utilization
Excellent C++ programming and software design skills
Proven experience developing, debugging, and optimizing high-performance applications, preferably with GPU acceleration using CUDA
Strong collaboration, communication, and documentation habits
Preferred
Experience developing or deploying Graph Neural Network solutions using PyTorch Geometric, or a similar framework
Experience working with data warehouse and lakehouse platforms, such as Snowflake or Databricks
Experience in two or more of the following domains: finance, cybersecurity, government or national laboratories, and retail
Strong understanding of system architecture, CPU, GPU, memory, and storage systems, as well as performance optimization
Experience with customer engagement and technical support, particularly for data science workflows and with vector search and storage solutions, such as FAISS or Milvus