Systems Engineer (Physical AI)

Location: San Francisco; in-person work 4 days per week in the SF Mission district office

Salary: $150K–$250K annually, plus equity

Company

hirify.global is a startup building Daft, an open-source distributed data engine and video-native indexing infrastructure for multimodal Physical AI datasets.

What you will do

• Design and build a video-native dataloader that uses rank-aware sampling, NVMe caching, random clip access, and direct tensor delivery to GPUs.
• Profile and optimize the data path from object storage through NVMe, page cache, host RAM, and device RAM.
• Optimize throughput for H100, B200, GB200, and NVL72 systems while preparing for Vera Rubin bandwidth requirements.
• Own performance benchmarks against customer and internal dataloader baselines, catching regressions at pull-request time.
• Partner with Physical AI researchers to integrate the loader into training stacks and measure end-to-end model utilization.
• Collaborate with Storage Infrastructure and Visual Understanding teams on index, format, and model-output ingestion interfaces.

Requirements

• Strong systems-level performance mindset and experience diagnosing bottlenecks with profiling tools and flamegraphs.
• Professional expertise in Rust, C++, or C.
• Strong operating-systems knowledge, including page cache, scheduling, syscalls, NUMA, and memory hierarchies.
• Understanding of data movement across NVMe, memory, network, PCIe, and NVLink, including throughput and latency constraints.
• Experience with SLURM, Kubernetes for GPU workloads, or another HPC scheduler.
• Ability to work in person 4 days per week from the San Francisco office.

Nice to have

• Hands-on CUDA and GPU experience.
• Expertise in memory and caching subsystems, including hugepages, NUMA pinning, and GPU-Direct Storage.
• Experience with video decoding pipelines such as PyAV, decord, or NVDEC, or with PyTorch DataLoader internals.
• Open-source contributions to systems projects in Rust or C++.

Culture & Benefits

• Small, tight-knit engineering team working closely with leading Physical AI labs and infrastructure companies.
• Competitive compensation and meaningful startup equity.
• Health, vision, and dental coverage.
• Flexible PTO, commuter benefits, and a 401(k) plan with employer match.
• Catered lunches and dinners for San Francisco employees, team-building events, and poker nights.
• Latest Apple equipment.