Low-Level Systems Engineer

Location: San Francisco, United States; onsite in person five days a week

Company

hirify.global is building a globally distributed inference cloud that runs customer requests on its own GPU fleet across multiple clouds using Kubernetes.

What you will do

• Write and tune low-level code running directly on GPUs.
• Increase hardware throughput and cost efficiency for inference workloads.
• Shrink models without reducing their quality.
• Investigate and resolve the causes of GPU underutilization.
• Profile systems before and after changes and validate performance measurements.

Requirements

• Around five or more years of low-level performance work where system speed was critical, including experience in operating systems, drivers, databases, compilers, trading systems, or inference engines.
• Strong C or C++ skills and genuine low-level expertise with memory, concurrency, caches, and hardware behavior.
• A proven track record of significantly improving performance, with the ability to explain the original bottleneck and the changes made.
• Strong measurement and profiling discipline, including before-and-after profiling and skepticism toward unreliable test setups.

Nice to have

• Operating system, hypervisor, driver, firmware, or real-time systems experience.
• Performance tools built for other engineers.
• Compiler development experience.
• GPU experience with NVIDIA, AMD, or open-source inference projects.

Culture & Benefits

• Small, venture-backed team with customers and revenue, offering substantial responsibility and room to grow.
• Work onsite in San Francisco five days a week.
• Competitive compensation plus equity.