PhD AI Agentic/ML System Co-op

What you'll do

• Work with state-of-the-art open-weight large language models to accelerate research and development.
• Leverage frontier coding agents to improve research and development workflows.
• Contribute to agentic system design and software optimizations for machine learning workloads.
• Optimize hardware performance for ML workloads across AMD GPUs, vLLM, SGLang, and PyTorch.
• Explore techniques to improve the efficiency, scalability, and performance of LLMs on modern hardware.
• Document research findings and contribute to technical reports, presentations, and research publications.

What we're looking for

• Currently enrolled in a PhD program in Computer Science, Electrical Engineering, or a related field at a U.S.-based university.
• PhD candidates are preferred.
• Proficiency in Python and/or C++ with exposure to performance profiling or optimization.
• General understanding of machine learning software stacks, large language models, and frameworks such as PyTorch or JAX.
• Demonstrated knowledge or experience in agentic software engineering, agent-harness development, or agent and skill evaluation.
• Demonstrated knowledge or experience in performance analysis, GPU architecture, GPU kernel development, or machine learning compilers.
• Ability to work full-time (40 hours a week) onsite or remote for the duration of the co-op term.
• PhD degree.