🛠 Tech Stack

💼 About This Role

You'll build automated performance and diagnostic tools for next-generation AI infrastructure at a leading AI inference platform. You'll measure GPU FLOPS, stress-test clusters, and define benchmarks that ensure production readiness. This role offers deep hardware exposure and high ownership.

🎯 What You'll Do

• Run and automate LLM benchmark suites like GSM8K and MMLU.
• Create automated acceptance tests for new GPU clusters.
• Develop internal GPU-enabled development environments.
• Build tools for automated model evaluation and optimization.

📋 Requirements

• Python familiarity
• Interest in GPU memory subsystems and networking
• Automation mindset; scripting repetitive tasks
• Desire to understand Transformer math and FLOPs

✨ Nice to Have

• C++ familiarity
• Experience with NVIDIA Nsight Systems or PyTorch Profiler
• Knowledge of quantization or speculative decoding

🎁 Benefits & Perks

• 💰 Competitive compensation with meaningful equity
• 🏥 100% medical, dental, vision for employee and dependents
• 🏖️ Flexible PTO including Winter Break
• 👶 Paid parental leave
• 🏦 Company-facilitated 401(k)