🛠 Tech Stack

💼 About This Role

You'll join a cutting-edge AI infrastructure team focused on powering the next generation of machine learning and large-scale AI workloads. You'll contribute directly to benchmarking and optimizing advanced GPU platforms that support training and inference for complex neural networks. This role offers the opportunity to work at the intersection of GPU performance engineering, deep learning optimization, and cloud-scale infrastructure development.

🎯 What You'll Do

• Benchmark GPU platform performance for ML and AI workloads across architectures.
• Profile GPU performance at system and kernel levels to identify optimizations.
• Analyze and improve efficiency, scalability, and utilization of training/inference workloads.
• Conduct acceptance testing for new GPU clusters to ensure operational readiness.
• Develop internal tools and dashboards for performance metrics and infrastructure trends.

📋 Requirements

• Strong foundation in machine learning and deep learning architectures.
• Deep understanding of performance optimization for large neural network training/inference.
• Extensive experience with modern deep learning frameworks such as PyTorch, JAX, or Megatron-LM.
• Solid expertise with GPU technologies including CUDA, NCCL, and performance libraries.
• Experience profiling GPU workloads using tools like Nsight or nvprof.

✨ Nice to Have

• Experience with LLM inference frameworks such as vLLM or SGLang.
• Familiarity with cloud-based ML ecosystems like AWS or GCP.
• Contributions to open-source ML tooling or benchmarking frameworks.

🎁 Benefits & Perks

• 💰 Competitive compensation aligned with experience.
• 🏠 Flexible remote work supporting work-life balance.
• 📚 Continuous learning and career development opportunities.
• 🌍 International work environment with globally distributed teams.
• 🚀 Impactful AI projects shaping the future of ML infrastructure.

📨 Hiring Process

Estimated timeline: 2-4 weeks · AI estimate

1. 1Recruiter call· 30 min
2. 2Technical interview· 60 min
3. 3Offer· 30 min