CUDA / AI Engineer

MWDN is seeking a CUDA / AI Engineer to tackle complex challenges at the intersection of AI systems and high-performance computing. The role involves investigating and optimizing LLM inference performance on modern GPUs, utilizing tools such as vLLM, SGLang, and TensorRT-LLM. Candidates should possess deep expertise in GPU computing, CUDA, and LLM inference optimization, with a strong understanding of GPU architecture, low-level kernels, and performance bottlenecks. Daily tasks will include analyzing GPU execution, memory access patterns, and kernel performance, as well as developing, optimizing, debugging, or reverse-engineering CUDA kernels. The position is suitable for individuals who enjoy delving into technical details and understanding system functionalities. While the employment type is not specified, MWDN emphasizes a stable and professional environment, supporting career development and legal assistance for independent contractors. Fluency in English is implied for effective communication in this international setting.