Research Engineer, Education

Research Engineers at Google DeepMind are the bridge between ambitious research ideas and functional, large-scale systems. They work in close collaboration with Research Scientists to design and build the infrastructure, models, and tools necessary to drive groundbreaking research.

For the AI for Education team, this means designing evaluations, building agentic systems and developing advanced reinforcement learning methodologies to tackle complex educational tasks. As a Research Engineer, you will be instrumental to transforming research hypotheses into tangible experiments. Your contributions will enable the team to make rapid, empirically-driven progress towards AI systems that can safely and effectively support learners everywhere. You will also have the opportunity to collaborate directly on projects with significant social impact.

Key responsibilities

• Design and build experiments and systems to rapidly test research hypotheses and demonstrate novel educational applications.
• Design, build, and own end-to-end infrastructure for long-term research in AI for education.
• Implement advanced machine learning methods to advance educational AI capabilities.
• Collaborate closely with research scientists, learning science experts, and product teams to translate research into robust experiments and land impactful research projects in real-world products.

• MSc in Computer Science (or higher) or a related technical field, or equivalent practical experience.
• Exceptional software engineering skills in Python or other programming language(s).
• Background or experience in machine learning, statistics, and/or data science methodologies.

• Experience with reinforcement learning, agentic systems, or user simulation methodologies.
• Demonstrated experience in education technology, learning science or other socially impactful applications of technology, such as in healthcare.
• Familiarity with prompting large AI models for structured data generation. 
• Demonstrated ability to translate ambiguous research questions into functional experimental setups.