🛠 Tech Stack

💼 About This Role

You'll design and scale the infrastructure behind Tinker to enable seamless model fine-tuning for internal teams and external customers. Your work will optimize GPU utilization, multi-tenant scheduling, and developer-friendly APIs, ensuring users can focus on research without infrastructure concerns. This role sits at the intersection of large-scale training systems and product infrastructure, with opportunities to publish and open-source your learnings.

🎯 What You'll Do

• Design and implement distributed job orchestration for multi-tenant workloads.
• Optimize GPU utilization, throughput, and reliability across clusters.
• Develop reusable frameworks to improve transparency, reproducibility, and performance.
• Co-design fine-tuning challenges into product features with researchers and engineers.

📋 Requirements

• Bachelor's degree or equivalent in computer science or related field.
• Understanding of deep learning frameworks like PyTorch or JAX.
• Thrive in a highly collaborative environment with cross-functional partners.
• Strong engineering skills with ability to debug complex codebases.

✨ Nice to Have

• Hands-on experience with container orchestration for GPU workloads.
• Background in multi-tenant platform design and storage systems for ML artifacts.
• Contributions to ML systems OSS like PyTorch/DeepSpeed/XLA.

🎁 Benefits & Perks

• 🏥 Health, dental, and vision insurance
• 🏖️ Unlimited PTO
• 👶 Paid parental leave
• 🚚 Relocation support
• 💰 Generous compensation

📨 Hiring Process

This is an evergreen role; applications reviewed continuously, and you may reapply after 6 months.

🚩 Heads Up

• Evergreen role with no immediate opening may lead to delayed response.
• High experience level expected given senior salary range.
• Vague requirement: 'thrive in a highly collaborative environment' lacks measurable criteria.