Multimodal AI Research Engineer

OpenAI is looking for a Multimodal AI Research Engineer to work on the next generation of models that understand and generate across text, images, audio, and video. You will develop novel architectures and training techniques for multimodal understanding. This role is at the frontier of AI research, working on models that can seamlessly reason across different modalities. You will help build the foundation for more capable and versatile AI systems. The ideal candidate has experience with vision-language models, audio models, or cross-modal learning.

- 5+ years of ML research/engineering experience - Experience with multimodal architectures (vision-language models, etc.) - Expert PyTorch and distributed training skills - Published research in multimodal AI - Experience with large-scale model training - PhD in ML/CS preferred