Research Intern, Frontier Agents (Winter 2027)

Together AI is looking for a research intern to join their team in San Francisco, working on the development of sophisticated AI systems that can perform multi-step tasks involving both text and speech. The focus is on creating models that can operate reliably in complex scientific and agentic scenarios. The intern will be involved in designing innovative training strategies, assembling datasets for specialized tasks, and analyzing failure points to enhance system robustness.

The position requires hands-on experimentation, applying cutting-edge techniques, and sharing insights through scientific papers and blog posts. Applicants should be enrolled in a Ph.D. program related to computer science or electrical engineering, with a solid grasp of machine learning principles, deep learning tools, and transformer-based models. The internship lasts approximately 12 to 14 weeks, offering competitive compensation, housing support, and opportunities to contribute to pioneering AI research.

The role can be performed remotely or onsite in San Francisco, with collaboration across various teams to develop scalable infrastructure and safety measures for advanced AI agents. Projects include training frontier models, creating datasets for alignment, and developing systems for complex multi-modal tasks. This internship is ideal for individuals eager to contribute to the evolution of autonomous AI systems and share their findings with the scientific community.