Research Intern, Inference (Summer 2027)

Together AI is seeking a research intern to join their Inference team, which focuses on developing efficient systems for large foundation models. The team works at the intersection of model architecture, systems engineering, and hardware optimization to improve AI serving performance. The intern will contribute to designing experiments, exploring distributed inference, and optimizing inference processes, working closely with cutting-edge AI technologies in a collaborative environment.

The role involves hands-on research in areas like compiler-aware optimization, large-scale serving architectures, and hardware-software co-design. Candidates should have a background in machine learning, deep learning frameworks, and programming in Python, with familiarity with Transformer models. The internship is remote or onsite in San Francisco, lasting 12 to 14 weeks, with competitive pay and benefits.

This internship is ideal for students pursuing advanced degrees in computer science or electrical engineering, especially those with research experience in foundation models or ML systems. Prior publication at major conferences or experience with CUDA programming and open-source contributions are preferred. The position offers a unique opportunity to work on scalable AI systems in a fast-growing company dedicated to open and transparent AI development.