Multimodal AI Research Fellow — Real-Time Avatar Tech

Nuance Labs; Seattle (WA); On-site; USD 200,000 - 350,000; Full time

Job summary

Nuance Labs is seeking a Research Fellow to work at the frontier of Multimodal LLMs, generative modeling, and real-time audiovisual AI. This 3-month fellowship is open to current PhD students and recent graduates from BS, MS, or PhD programs.

As a fellow, you’ll own a concrete research problem, train and evaluate models, and collaborate with senior researchers to translate work into production-grade experiments. A strong publication record is valued.

Qualifications

• Strong working knowledge of PyTorch and deep learning — you can train a model, debug a training run, and reason about loss.
• At least one first-author paper at a tier 1 venue — or equivalent evidence of strong research taste and execution.
• Genuine interest in joining Nuance full-time after the fellowship. We are looking for long-term partners on this journey.

Responsibilities

• Own a concrete research problem from framing through experiments, analysis, and integration into the Nuance stack.
• Work on frontier Multimodal LLM systems spanning audio, video, language, and real-time interaction.
• Explore and adapt modern generative modeling techniques, including flow matching, diffusion, autoregressive modeling, and hybrid approaches.
• Read papers, reproduce key results, and turn promising ideas into production-grade experiments.
• Design, instrument, debug, and interpret training and evaluation runs with scientific rigor.

Skills

PyTorch

Deep learning

First-author paper

Education

BS/MS/PhD in CS/ML/Math/Physics/EE

Tools

PyTorch

TensorFlow

ML frameworks

Job description

Nuance Labs is seeking a Research Fellow to work at the frontier of Multimodal LLMs, generative modeling, and real-time audiovisual AI. This 3-month fellowship is open to current PhD students and recent graduates from BS, MS, or PhD programs.

As a fellow, you’ll own a concrete research problem, train and evaluate models, and collaborate with senior researchers to translate work into production-grade experiments. A strong publication record is valued.