Research Engineer, Multimodal AI

You will be part of the REMY (Research & dEvelopment in Multimodal technologY) team in the Media Understanding organization, at Google DeepMind. In this role, you will have the opportunity to push forward state-of-the-art research in multimodal AI representation models, in the context of recent advancements in multimodal foundation models generally. You'll be at the forefront of developing models that power Google products used by billions of people worldwide. Your work will directly impact how these products understand and interact with images, text and video. This is a unique opportunity to shape the future of multimodal AI and its applications in a dynamic and impactful environment.

We are a team of research/software engineers, research scientists, and machine learning experts, working together to enable superhuman understanding of the multimodal world.

You'll be developing the next SOTA models for multimodal understanding. Your work will include researching new modeling techniques, implementing research ideas, running experiments to evaluate improvements, and identifying new opportunities.

Key responsibilities:

As a member of the Media Understanding team, you will be responsible for conducting fundamental and applied research in multimodal AI (computer vision, language understanding, machine learning, and related areas). Your job responsibilities will include:

• Conducting core research in the areas of computer vision, language understanding, multimodal models, large scale AI models and other key computer vision tasks.
• Training and evaluating AI models for a variety of use cases. 
• Researching, implementing, and adapting state of the art deep learning approaches for Google’s use cases
• Collaborating closely with other GDM and partner teams to make progress towards building the most advanced representation models.