🛠 Tech Stack

💼 About This Role

You'll join an elite AI startup to design and improve training infrastructure for large-scale models. Your core impact will be optimizing GPU performance for deep learning workloads using CUDA and PyTorch. You'll also contribute to post-training processes like reinforcement learning and model serving.

🎯 What You'll Do

• Implement improvements to training infrastructure for large-scale models
• Optimize GPU kernel performance using CUDA and profiling tools
• Scale GPU jobs across compute clusters using Slurm or Kubernetes
• Contribute to post-training processes including reinforcement learning

📋 Requirements

• Python and PyTorch fluency
• Experience implementing and training large deep learning models
• Proficiency in low-level GPU code (CUDA, C++)
• 3+ years scaling GPU jobs on large clusters (Slurm or Kubernetes)

✨ Nice to Have

• Experience with model serving infrastructure
• Background in reinforcement learning or fine-tuning

🎁 Benefits & Perks

• 🏖️ Five weeks paid leave
• 🏥 Comprehensive healthcare including vision and dental
• 💻 Remote-first culture with global team
• 🛂 Visa support including H1B and OPT transfers