Research Scientist in Speech Foundation Model - Seed - Graduates - 2027 Start (PhD)

Join the Seed Speech team to develop and scale speech foundation models for understanding and generation. You’ll design training pipelines (data construction, instruction tuning, and model alignment), improve capabilities like speech recognition, synthesis, reasoning, and robustness, and optimize architectures for training efficiency and system performance. The role also includes exploring natural, interactive interfaces for speech-based systems as part of cutting-edge speech, audio, and multimodal deep learning research.