Intern, Systems Software Engineering

What you'll do

• Profile the RL post-training loop to identify memory limits, data-transfer overhead, and GPU idle time.
• Develop a runtime feature to coordinate bounded memory staging, asynchronous storage writes, and capacity allocation.
• Integrate the developed feature with an existing RL framework using training and checkpoint interfaces.
• Preserve snapshot consistency, policy-version provenance, and durable checkpoint publication during asynchronous operations.
• Compare memory-only, storage-only, and combined pool configurations using throughput and restart performance metrics.
• Deliver correctness tests, reproducible experiments, design documentation, and code suitable for open-source contribution.
• Identify and resolve resource management issues on post-RL workloads within the host OS.

What we're looking for

• Must be pursuing a Master's or PhD in Computer Science.
• Must have at least one academic quarter or semester remaining.
• Must have programming experience in C, C++, or Rust.
• Must have experience with Python or shell scripting for experiments and automation.
• Must have an understanding of operating systems, memory management, concurrency, and computer architecture.
• Must have experience developing and debugging on Linux, using Git, and communicating technical results.
• Experience with CXL, NUMA, Device-DAX, mmap, shared-memory programming, or memory allocators (preferred).
• Experience with GPU programming, CUDA, DMA, or profiling memory and data-transfer performance (preferred).