Multimodal AI Researcher

What you'll do

• Develop foundation models for generative AI and multimodal systems integrating video, audio, and text.
• Research and develop algorithms for interactive models and audio-to-audio modeling systems.
• Solve open research problems in multimodal generative AI and intelligent agents.
• Define data requirements, validation strategies, and key performance indicators for product features.
• Train and tune large-scale foundation models using architectures like diffusion, reinforcement learning, or flow matching.
• Develop real-time or streaming multimodal models for integration into consumer products.
• Implement research findings through practical software engineering using Python and PyTorch.

What we're looking for

• Bachelor's degree and at least 3 years of relevant industry experience.
• Experience building models for multimodal perception systems.
• Experience working with Large Language Models (LLMs) and Vision Language Models (VLMs).
• Proficiency in Python, PyTorch, and software engineering skills.
• MS or PhD in computer vision, graphics, machine learning, computer science, or related fields (preferred).
• Experience developing, training, or tuning foundation models and multimodal LLMs (preferred).
• Experience with generative architectures like diffusion, reinforcement learning, flow matching, or normalizing flow at scale (preferred).
• Experience with real-time/streaming multimodal models, speech understanding, or applying RL to post-train foundation models (preferred).