🛠 Tech Stack

💼 About This Role

You'll design and train large-scale multimodal generative models for audio generation at Mirelo AI. Your work will directly shape foundational AI models that "unmute" silent video content. This role offers deep research autonomy backed by $41M in seed funding.

🎯 What You'll Do

• Design and train large-scale multimodal generative models for audio.
• Explore novel modeling ideas for music, sound, and speech generation.
• Develop post-training techniques for fine-grained control and editing.
• Conduct rigorous ablation studies and communicate actionable insights.

📋 Requirements

• Hands-on experience training large-scale generative models in a fast-paced research environment.
• Deep understanding of ML research in image, language, video, or audio.
• Strong proficiency in PyTorch and transformer architectures.
• Solid understanding of distributed training techniques (FSDP, low precision, model parallelism).

✨ Nice to Have

• Proficiency with profiling and debugging multi-GPU operations using Nsight.
• Strong software engineering skills in large codebases beyond research.
• Experience with generative models for audio and audio codec design.

🎁 Benefits & Perks

• 💰 Competitive compensation and equity sharing in success.
• 🚀 True ownership from day one with genuine autonomy.
• 🎯 Build for the next generation of creators.
• 🌍 Join at a pivotal moment with fresh $41M funding.