Summer 2026 ML Engineer/Research Intern

Languages: Python (PyTorch), JavaScript (TypeScript)

Listen and try Sonauto: https://sonauto.ai/

About the role:

We're training state of the art generative models in a completely new modality where the standard of quality is significantly higher than those that came before. Accordingly, we are consistently implementing new generative architectures and improving current ones. Previous interns’ experiments were in production within weeks and their names were front and center in our announcements.

Right now we're doing distributed training at the scale of hundreds of H100s on diffusion models, GANs, language models, and more written in PyTorch.

You will lead research efforts for improving the song and audio quality of our generative music models, along with productizing the results.

We’re looking for people with:

• Extensive experience with PyTorch training generative models like diffusion models, GANs, and language models. Work with open-source orgs/projects counts too.

• Experience with JavaScript (our frontend is React/NextJS) and backend development are also important to us.

• Experience with audio models is a huge plus.

• Many of our biggest improvements wouldn’t have been possible without extremely close listening, so being an audiophile and/or having music production experience are also huge pluses.