GPU AI Engineer: CUDA & LLM Inference Optimizer

MWDN Ltd.; Lisboa; On-site; EUR 70,000 - 110,000; Full time

Benefits offered by this job

Flexible working hours

29 days of PTO (18 working days per yr

10 paid recovery days

Full financial and legal support for 0

Free English classes

Dedicated HR support

Team-building events

Friendly company culture

Job summary

MWDN Ltd. is seeking a highly technical engineer or researcher with deep CUDA and GPU expertise to optimize LLM inference on modern GPUs. You will work on AI systems at the intersection of HPC and high-performance computing, collaborating with global teams.

Ideal candidates will have hands-on CUDA experience, knowledge of GPU architecture, and familiarity with vLLM/SGLang/TensorRT-LLM/Triton/FlashAttention. The role offers flexible hours and full support for independent contractors.

Qualifications

• Strong hands-on experience with CUDA and GPU programming.
• Deep understanding of GPU memory hierarchy and performance optimization.
• Experience with LLM inference, AI systems, or HPC.
• English communication skills.

Responsibilities

• Investigate and optimize LLM inference performance on modern GPUs.
• Work with inference frameworks such as vLLM, SGLang, TensorRT-LLM, or similar systems.
• Analyze GPU execution, memory access patterns, and kernel performance.
• Develop, optimize, and debug CUDA kernels and low-level GPU operations.
• Explore optimization techniques for matrix multiply and attention mechanisms.
• Benchmark latency, throughput, and GPU utilization and propose improvements.

Skills

CUDA programming

GPU architecture

AI systems

Analytical thinking

Tools

vLLM

SGLang

TensorRT-LLM

Triton

FlashAttention

Job description

MWDN Ltd. is seeking a highly technical engineer or researcher with deep CUDA and GPU expertise to optimize LLM inference on modern GPUs. You will work on AI systems at the intersection of HPC and high-performance computing, collaborating with global teams.

Ideal candidates will have hands-on CUDA experience, knowledge of GPU architecture, and familiarity with vLLM/SGLang/TensorRT-LLM/Triton/FlashAttention. The role offers flexible hours and full support for independent contractors.