AI Research Scientist, Multimodal - Monetization GenAI

Meta; Menlo Park (CA); On-site; USD 180,000 - 280,000; Full time

Job summary

Meta's Monetization GenAI Video team is building the next generation of generative AI for video—creating, understanding, and transforming at scale to reach billions of users. We develop foundation models for video generation, video-to-video editing, audio generation, and multimodal video understanding powered by large language models.

The team seeks AI Research Scientists to define the research vision across video and audio generation, drive breakthroughs, and shape the technical direction of

Qualifications

• Bachelor's degree in CS/engineering or related field.
• PhD in CS/AI/ML or related field.
• Experience as technical lead on a team and/or managing complex projects end-to-end.
• 2+ years of training large language and/or vision models with multimodal LLMs.
• Research expertise in video generation/understanding, multimodal learning, or diffusion models.
• Notable industry research contributions with top conferences (e.g., ACL, NeurIPS, ICML, ICLR, AAAI, CVPR).
• Experience with frontier-quality/state-of-the-art MLLMs.
• Experience with audio/speech generation or processing.
• Experience with unified/multi-task foundation model architectures.
• Industry research lab experience.

Responsibilities

• Lead end-to-end AI research and model development for video-centric generative AI across Meta's advertising surfaces.
• Drive advancements in video generation & enhancement.
• Develop video-to-video & audio generation capabilities.
• Advance video & visual understanding through novel research.
• Conduct foundation model research to support generative AI innovation.
• Define research agendas and pioneer new directions in video/audio generation and multimodal understanding.

Skills

Leadership experience

LLM training

Multimodal learning

Video generation/understanding

Diffusion models

Industry recognition

Education

Bachelor's degree in Computer Science, Computer Engineering or related field

PhD in Computer Science/AI/ML or related field

Job description

About

The Monetization GenAI Video team, part of the Ads pillar, is building the next generation of generative AI for video - from creation to understanding to transformation - at a scale that reaches billions of users. We develop proprietary foundation models for video generation, video-to-video editing, audio generation (music, voiceover, dubbing), and multimodal video understanding powered by large language models. We produce industry leading video models, and our research translates into products that drive hundreds of millions of dollars in annual advertising revenue. We are looking for AI Research Scientists to define the research vision across video and audio generation, drive breakthrough research, and shape the technical direction of the organization.

Responsibilities

• Lead end-to-end AI research and model development for video-centric generative AI across Meta's advertising surfaces.
• Drive advancements in video generation & enhancement.
• Develop video-to-video & audio generation capabilities.
• Advance video & visual understanding through novel research.
• Conduct foundation model research to support generative AI innovation.
• Define research agendas and pioneer new directions in video/audio generation and multimodal understanding.

Minimum Qualifications

• Bachelor's degree in Computer Science, Computer Engineering, relevant technical field, or equivalent practical experience
• Has obtained a PhD in Computer Science, AI/ML, or a relevant technical field
• Experience as a technical lead on a team and/or leading multiple complex technical projects from end-to-end
• 2+ years of experience training large language and/or vision models, with extensive and recent experience training multimodal LLMs
• Research expertise in video generation/understanding, multimodal learning, or diffusion models
• Demonstrated significant industry influence in the field of AI and/or recently published research in leading peer-reviewed conferences (e.g., ACL, NeurIPS, ICML, ICLR, AAAI, KDD, CVPR, ICCV) First-author publications at top peer-reviewed conferences (e.g., ACL, NeurIPS, ICML, ICLR, AAAI, KDD, CVPR, ICCV)
• Experience working on frontier-quality/state-of-the-art MLLM
• Experience with audio/speech generation or processing
• Experience with unified/multi-task foundation model architectures
• Industry research lab experience