Research Engineer, Gemini Post Training, Model Behavior
About The Position
Snapshot Artificial Intelligence could be one of humanity’s most useful inventions. At DeepMind, we’re a team of scientists, engineers, machine learning experts and more, working together to advance the state of the art in artificial intelligence. We use our technologies for widespread public benefit and scientific discovery, and collaborate with others on critical challenges, ensuring safety and ethics are the highest priority. About Us The Model Behavior team focuses on aligning Gemini models with human needs. We ensure that Gemini is safe, robust, and reliable, while optimizing its helpfulness to be context-aware and deeply aligned with user intent. The Role We are looking for a Research Engineer with proven leadership experience building large-scale machine learning systems, with expertise in modeling, data engineering and evaluations, to help us develop the best models as part of the Gemini Post-Training team, with a focus on model behavior. In this role, you will help execute the team’s strategic roadmap, identify critical modeling needs and collaborate cross-functionally and across the stack to land improvements that truly make a difference to the user.
Requirements
- BS, MS or Ph.D. in Computer Science, Artificial Intelligence, or a related field, or equivalent practical experience
- Fluency in Python or C++ and JAX or similar ML framework
- Software engineering, ML and data engineering experience (e.g. model training / deployment, efficiency optimization, data pipeline design, ML infra)
- Experience building robust automatic or human evaluations for ML models
- Proven ability to lead engineering teams in fast-moving environments, under ambiguity
- Excellent communication and cross-functional collaboration skills, willingness to unblock others and pick up slack
Nice To Haves
- Have hands-on experience with training and debugging LLMs.
- Experience with RL fine tuning methods are a plus.
- Enjoy the end-to-end process of productionizing cutting-edge research and taking projects from proof-of-concept to launch are passionate about applying AI for learning use cases
- obsess over the user and are excited about incorporating user feedback into the model development process
- Are well versed in Google infrastructure, including for fine-tuning and serving Gemini models
Responsibilities
- Develop techniques and data pipelines to optimize frontier Gemini models on key capabilities important for model behavior
- Build robust and real-world motivated evaluations
- Partner with feature teams to prioritize modeling needs and ship gemini powered features.
- Establish best practices for code health and documentation on the team, to facilitate collaboration and reliable development.
- Mentor and guide junior engineers, fostering their professional growth and development.
Stand Out From the Crowd
Upload your resume and get instant feedback on how well it matches this job.
What This Job Offers
Job Type
Full-time
Career Level
Mid Level
Education Level
Ph.D. or professional degree
