AI Systems Engineer – GPU/ROCm/CUDA | ML Frameworks Optimization

Job Title: AI Systems Engineer – GPU/ROCm/CUDA | ML Frameworks Optimization

• Location: Hyderabad
• Experience : 3-6 [Mid-Senior]

Job Description:

We are looking for a passionate and experienced AI Systems Engineer to join our team to work on next-generation Machine Learning technologies and optimize performance across AMD GPU accelerators. This role involves low-level GPU programming, custom ML kernel development, and working with state-of-the-art inference engines.

Key Responsibilities:

• Develop and optimize custom Deep Learning GPU kernels using ROCm/CUDA or shader languages
• Support and enhance ML model deployment on Linux platforms
• Optimize performance of ROCm drivers and inferencing engines for AI/ML workloads
• Collaborate closely with internal hardware/software teams to support next-gen GPU accelerators
• Profile, debug, and improve performance of GPU kernels and AI model pipelines
• Contribute to designing and implementing new AI technologies and workflows

Required Skills & Qualifications:

• BS/MS in Computer Science, Electrical Engineering, or equivalent
• Strong programming skills in C/C++, Python
• Solid experience working with Linux CLI, bash scripting, or PowerShell
• Hands-on experience with Python ML libraries such as PyTorch, Transformers
• Knowledge of writing high-performance ML kernels using Triton, JAX, or similar
• Experience with debugging tools like gdb, valgrind, and profiling tools such as nsys, rocprof
• Familiarity with AI inferencing runtimes such as vllm, ollama, llama.cpp, or sglang
• Understanding of GPU and PC architecture, x86/x64 instruction sets
• Experience developing with ROCm, CUDA, or shader programming

Nice to Have:

• Knowledge of x86 Assembly
• Contributions to open-source ML/DL performance libraries
• Exposure to compiler optimization techniques for GPU code

What We Offer:

• Work on cutting-edge GPU technologies and ML systems
• Exposure to performance-critical AI workloads
• Collaborative and research-oriented environment
• Competitive compensation and career growth opportunities

Apply: If you are looking for job change share your updated resume to vagdevi@semi-leaf.com

Job Type: Full-time

Pay: Up to ₹3,000,000.00 per year

Experience:

• Deep Learning GPU kernels using ROCm/CUDA: 2 years (Required)
• programming skills in C/C++, Python: 1 year (Required)
• Python ML libraries such as PyTorch, Transformers: 1 year (Required)
• developing with ROCm, CUDA, : 1 year (Required)

Work Location: In person

Speak with the employer
+91 7483459258