ML Engineer / Researcher (Generative AI Music)
📍 San Francisco, CA | 🕐 Full-Time, On-Site | 💰 $120K – $230K + Equity
About the Company
Our client is building the next generation of AI-powered music creation. Their platform allows users to turn prompts, lyrics, and melodies into full songs using cutting-edge generative models. With a product-first mindset and a focus on audio quality, they're shipping fast, innovating constantly, and enabling creative expression like never before.
About the Role
As an ML Engineer / Researcher, you’ll lead efforts to improve the audio and song quality of generative music models. You'll train and implement state-of-the-art architectures (diffusion models, GANs, LLMs) and ship them directly into a live product used by thousands.
This is not a pure research role — your models will go into production rapidly, often before training is fully complete. You'll also contribute to backend and frontend development where needed.
Responsibilities
- Train and improve generative models using PyTorch
- Scale distributed training across multi-GPU infrastructure (H100s)
- Collaborate with engineering to deploy models to production
- Conduct research to improve generative audio quality
- Help build product features powered by your models
Requirements
- 1+ years of experience training diffusion models, GANs, or LLMs in PyTorch
- Experience with distributed training (multi-GPU or multi-node)
- Experience with JavaScript (React/Next.js) and backend development
- Bonus: Experience with audio models or music production
- Bonus: Audiophile-level attention to sound quality
