GPU Compute Performance Engineer | Scalable AI Systems

Tangier Partners; Boston (MA); On-site; USD 150,000 - 210,000; Full time

Job summary

Tangier Partners is partnering with a well-funded AI research company to hire a Performance Engineer focused on scalable compute efficiency. You will own how compute infrastructure runs at scale, profile GPU workloads, and drive capacity planning from the ground up.

You will shape performance strategies across research and production, collaborating with engineers to translate workload patterns into robust infrastructure requirements and tooling for long-term improvement.

Qualifications

• Hands-on experience optimizing GPU workloads and distributed computing environments.
• Strong grasp of systems performance profiling, benchmarking, and capacity planning.
• Experience with large-scale compute infrastructure in research or AI environments.
• Ability to operate with high autonomy and bring structure to a greenfield function.
• Comfort at the intersection of research and engineering.

Responsibilities

• Profile and optimize GPU and compute-intensive workloads across research and production systems.
• Build capacity models to forecast compute needs as the business scales.
• Identify performance bottlenecks across distributed systems and drive resolution.
• Partner with research and engineering teams to understand workload patterns and translate them into infrastructure requirements.
• Define the frameworks and tooling the team will use to measure and improve system performance over time.

Skills

GPU optimization

Distributed computing

Systems performance profiling

Benchmarking

Capacity planning

Autonomy

Research & engineering collaboration

Job description

Tangier Partners is partnering with a well-funded AI research company to hire a Performance Engineer focused on scalable compute efficiency. You will own how compute infrastructure runs at scale, profile GPU workloads, and drive capacity planning from the ground up.

You will shape performance strategies across research and production, collaborating with engineers to translate workload patterns into robust infrastructure requirements and tooling for long-term improvement.