As a staff software engineer for GenAI Performance and Kernel, you will be responsible for the design, implementation, optimization, and correctness of high-performance GPU kernels for the GenAI inference stack. You will lead the development of compute paths while managing trade-offs between hardware efficiency and generality. This role involves close collaboration with ML researchers, systems engineers, and product teams to enhance inference performance at scale. You will also mentor other engineers in kernel-level performance engineering.

• BS/MS/PhD in Computer Science or a related field
• Deep hands-on experience writing and tuning compute kernels for ML workloads
• Strong knowledge of GPU/accelerator architecture
• Experience with advanced optimization techniques
• Familiarity with ML-specific kernel libraries or open kernels
• Strong debugging and profiling skills
• Experience reasoning about numerical stability and error propagation
• Experience integrating optimized kernels into real-world ML inference systems
• Experience building high-performance products leveraging GPU acceleration
• Excellent communication and leadership skills
• A track record of shipping performance-critical, high-quality production software
• Bonus: published in systems/ML performance venues or experience with custom accelerators

• Lead the design, implementation, benchmarking, and maintenance of core compute kernels optimized for various hardware backends.
• Drive the performance roadmap for kernel-level improvements including vectorization and auto-tuning.
• Integrate kernel optimizations with higher-level ML systems.
• Build and maintain profiling, instrumentation, and verification tooling.
• Lead performance investigations and root-cause analysis on inference bottlenecks.
• Establish coding patterns and frameworks to modularize kernels for reuse.
• Influence system architecture decisions to enhance kernel improvements.
• Mentor and guide other engineers working on lower-level performance.
• Collaborate with infrastructure, tooling, and ML teams to roll out kernel-level optimizations.