about 3 hours ago
Base Salary
$150k - $250k/yr
Responsibilities
- Design and build the video-native dataloader for GPUs.
- Profile and optimize the data path from object store to device RAM.
- Saturate the latest hardware on real customer training jobs.
- Own performance benchmarks against customer baselines.
- Partner with researchers to integrate the loader into their training stack.
- Collaborate with cross-functional teams on data ingestion paths.
Requirements
- Obsession with systems-level performance.
- Strong opinions on io_uring.
- Proficiency in Rust, C++, or C.
- Familiarity with operating systems and memory hierarchies.
- Understanding of data throughput and latency budgets.
Benefits
- In-person, tight-knit team working 4 days/week in SF.
- Competitive compensation and meaningful startup equity.
- Catered lunches and dinners for employees.
- Commuter benefits and team-building events.
- Health, vision, and dental coverage.
- Flexible PTO and latest Apple equipment.
- 401(k) plan with match.