About the role
At Boson AI, the engineering team develops advanced real-time speech and multimodal AI technologies that enable natural conversations with AI. Their products include speech-to-speech APIs supporting over 100 languages, expressive text-to-speech models with voice cloning, and AI-driven video avatars. The team consists of over 40 researchers and engineers working on cutting-edge models trained on a large GPU cluster, focusing on making AI interactions more human-like and seamless.
This internship is designed exclusively for Master of Science in Applied Computing (MScAC) students at the University of Toronto, offering an 8-month applied research experience starting in May 2027. Interns will collaborate with mentors on projects that directly impact Boson AI's models and products. The role is hybrid, based in Toronto, and involves hands-on work in areas such as audio data annotation, speech evaluation, video generation, efficient training, real-time inference, conversational agents, and full-duplex conversation modeling.
Candidates should have strong coding skills and relevant backgrounds depending on the project, such as machine learning, statistics, computer vision, systems optimization, or reinforcement learning. The internship suits those interested in scripting, experimentation, large datasets, GPU programming, dialog systems, or human-computer interaction. This opportunity provides practical experience in a fast-paced environment focused on speech and multimodal AI technologies.