Research Engineer Intern

Three-month research engineering internship on a production decentralized training system. The intern owns a scoped project from the live roadmap, ships real work early, and contributes daily to the production codebase. Responsibilities include building and improving multiprocessing, async I/O, and threading components in a distributed training stack, and working hands-on with large-scale training infrastructure across cloud providers. Requires current or recent Masters or PhD enrollment, strong Python and PyTorch, concurrent systems experience, and exposure to distributed ML.