About the job
Meta is seeking AI Research Scientists to join the Realtime AI Voice team in Meta Super-intelligence Labs. The team works on speech-LLMs, and leverages knowledge in areas like speech encoders, pre-training, post-training, reinforcement learning, LLM alignment, multimodal modeling, speech and audio processing, speech recognition, speech synthesis, and multilingual modeling.
Responsibilities
Lead, collaborate, and execute on research that pushes forward the state of the art in speech and large language model research
Directly contribute to experiments, including designing experimental details, develop reusable code, running evaluations, and organizing results
Help identify long-term research goals as well as intermediate milestones
Work cross-functionally to translate research breakthroughs into scalable, production-ready solutions for Meta's conversational AI / product experiences
Qualifications
Minimum
Bachelor's degree in Computer Science, Computer Engineering, relevant technical field, or equivalent practical experience
PhD degree in Computer Science, Mathematics, or similar quantitative field
2+ years of post-PhD experience in an academic, industry, or government laboratory setting, with primary responsibilities focused on AI research
Proven track record of publications at peer-reviewed AI & speech conferences (e.g. NeurIPS, ICML, ICLR, ICASSP)
Experience in training, fine-tuning, and/or experimenting with foundation models beyond black-box use
Familiarity with one or more deep learning frameworks (e.g., pytorch, tensorflow)
Experience communicating complex research to public audiences of peers
Preferred
High impact publications at leading AI and speech conferences and journals (e.g. e.g., NeurIPS, ICML, ICLR, ICASSP, Interspeech)
Experience with interdisciplinary and/or cross-collaboration
Experience bringing machine learning-based products from research to production