Research Engineer, Gemini Omni, DeepMind

Overview

In this role you apply research ideas to real-world problems by prototyping, curating data, training models, and deploying solutions at scale. You will advance Generative Media techniques (image, video, audio) and optimize inference while strengthening multimodal understanding. You collaborate across teams to test and deploy promising ideas, contributing to Google’s AI research ecosystem and broader product impact. This is a hands-on, research-driven position with a focus on scalable, ethical AI innovation.

Responsibilities

• Prototype ideas and build proof-of-concept implementations
• Curate datasets and train models for generative media and multimodal tasks
• Optimize algorithms and models for efficient inference and scaling
• Advance capabilities in multimodal understanding and cross-domain AI challenges
• Collaborate with cross-functional teams to deploy research outcomes
• Conduct applied research and contribute to iterative experimentation and deployment

Key requirements

• Master’s degree in computer science
• Experience with TensorFlow or ML frameworks (e.g. JAX or PyTorch)
• Industry experience delivering projects from proof-of-concept to implementation
• Experience with training diffusion models
• Experience conducting applied research

• cross-functional collaboration
• strong communication
• problem-solving and adaptability
• TensorFlow
• JAX
• PyTorch

…