🛠 Tech Stack

💼 About This Role

You'll join Snowflake's AI Research team to build efficient, scalable LLM inference and training systems. Your work will power agentic enterprise applications and reduce costs through innovations like SwiftKV. You'll collaborate with founding members of DeepSpeed, vLLM, and TensorFlow.

🎯 What You'll Do

• Analyze and optimize GPU kernel performance for LLM training and inference.
• Profile and benchmark deep learning systems to identify bottlenecks.
• Design optimizations to reduce latency and improve resource utilization.
• Contribute to agentic frameworks and applications for LLM workflows.

📋 Requirements

• Bachelor's degree in CS, EE, or related field (Master's or PhD preferred).
• 5 years of experience in GPU kernel, deep learning, or HPC optimization.
• Proficiency in deep learning frameworks such as PyTorch, TensorFlow, JAX.
• Experience with CUDA and GPU architectures.

✨ Nice to Have

• Experience with CUTLASS, Triton, cuDNN.
• Experience with profiling tools (nvprof, Nsight).

🎁 Benefits & Perks

• 💰 Competitive salary and equity.
• 🏠 Hybrid work flexibility.
• 📚 Innovation-driven culture with top-tier collaborators.
• 🌟 Opportunities to publish at top conferences.

📨 Hiring Process

Estimated timeline: 2-4 weeks · AI estimate

1. 1Recruiter Screen· 30 min
2. 2Technical Interview· 60 min
3. 3Onsite Interviews· 4 hours