Multimodal AI Engineer

Build image + text models powering our creative tools — generation, editing, and retrieval.

• Train and serve diffusion + multimodal models
• Own quality + safety evals for generation
• Ship inference that feels instant

• Diffusion / multimodal experience
• PyTorch fluency
• Care about output quality

Glasswing builds creative tooling for studios. 18 people, design-led, remote.