Senior Software Engineer, GNN
Core
Building large-scale machine learning solutions, specifically Graph Neural Networks (GNNs), Tabular Foundation Models, and ensemble models, with a focus on efficient training and inference on GPU infrastructure.
Role type
Senior IC software engineer (GPU-accelerated ML frameworks)
Builds
PyTorch-based frameworks, CUDA-X Libraries, and user-facing tools for NVIDIA's data science solutions
Domain
High-performance computing, GPU acceleration, machine learning infrastructure
Deliverable
production ML models
Required skills
C++ programming, PyTorch development, GPU acceleration (CUDA), distributed system training, high-performance application optimization, system architecture design, code refactoring, agentic coding tools
Preferred skills
PyTorch Geometric, data warehouse/lakehouse platforms (Snowflake, Databricks), vector search/storage (FAISS, Milvus), domain expertise in finance/cybersecurity/government/retail
Technologies
PyTorch, CUDA, CUDA-X Libraries, PyTorch Geometric, Snowflake, Databricks, FAISS, Milvus
Responsibilities
Develop accelerated PyTorch solutions for GNNs and ensemble models; support CUDA-X Libraries; partner with scientists to develop GPU-accelerated implementations; gather technical requirements from customers; provide technical leadership and mentorship; re-architecture codebase to reduce maintenance overhead; apply agentic coding tools for bug fixing and feature implementation
Seniority
Senior, hands-on IC with mentorship responsibilities
