Applied AI Scientist, Multimodal Intelligence

What you'll do

• Conduct applied research experiments to bridge the gap between research advances and practical generative AI applications.
• Evaluate and adapt emerging research in multimodal foundation models for product integration.
• Develop robust solutions by transforming promising research approaches into production-ready features.
• Manage large-scale data pipelines including curation, preprocessing, and efficient storage.
• Fine-tune and optimize large foundation models for deployment on various hardware platforms.
• Build autonomous or semi-autonomous agents using planning, task decomposition, and multi-step reasoning.
• Implement privacy-preserving or on-device machine learning techniques to meet product requirements.
• Develop innovative features that leverage multiple sensors like face tracking and vision systems.

What we're looking for

• Must have an MS in Computer Science, Computer Vision, Machine Learning, or a related technical field.
• Must have at least 6 years of relevant experience.
• A PhD or equivalent practical experience in a related technical field is preferred.
• Must have experience in deep learning with work in multimodal systems such as vision, language, or video.
• Must be proficient in Python and modern deep learning frameworks like PyTorch or JAX.
• Must have experience with rapid prototyping, reproduction, and validation of research ideas.
• Preferred experience includes a publication record in relevant conferences or a track record of applying deep learning to real-world products.
• Preferred experience includes foundation models, large-scale data pipelines, model optimization for production, and privacy-preserving or on-device machine learning.