Developer

Role Overview

GPU Kernel Developer (AI/ML) working on high-performance optimization and scaling of large language models (LLMs), with a focus on BERT and transformer workloads. The role is remote and full-time, centered on developing and tuning GPU kernels for modern ML frameworks.

Responsibilities

• Design, implement, and optimize GPU kernels for AI/ML workloads
• Concentrate on GEMM (matrix multiplication) and other performance-critical operations
• Collaborate with researchers and engineers to accelerate BERT and transformer-based models
• Integrate optimized kernels into PyTorch and Triton
• Perform profiling, benchmarking, and tuning across diverse GPU hardware/architectures
• Work with system engineers to support production deployment

Requirements (Must-Have)

• Strong GPU kernel development expertise
• Hands-on experience with PyTorch and Triton (mandatory)
• Proficiency in Python, C, and C++
• Deep understanding of system-level programming and GPU architecture
• Experience in AI/ML model optimization, especially BERT
• Strong knowledge of HPC and parallel programming

Nice-to-Haves

• Experience deploying large-scale AI/ML models
• Familiarity with CUDA (or other GPU programming frameworks)
• Background in distributed training systems
• Strong problem-solving and debugging skills