Post-Training Agentic Research Scientist - Generative AI

Google; New York (NY); On-site; USD 174,000 - 252,000; Full time

Benefits offered by this job

Equity

Bonus target

Benefits

Job summary

Google DeepMind is hiring a Post-training Agentic Research Scientist to advance Gemini and deploy agents at scale. You will identify model improvements, train and evaluate agents across diverse digital environments, and focus on multi-step reinforcement learning in multimodal settings to accelerate real-world deployments.

You will contribute to the broader research community through publication and collaboration, joining a mission to push AI capabilities while ensuring safe and ethical

Qualifications

• PhD or equivalent in CS/AI or related field.
• 2 years of experience developing and training ML models.
• Experience deploying models using deep learning frameworks (TensorFlow, PyTorch, JAX).
• Experience with Large Language Models, NLP, or Generative AI.
• Experience in Python programming.

Responsibilities

• Make core contributions to Gemini that advance what digital agents can do and are deployable into production.
• Identify model areas of improvement.
• Train and evaluate agents across various digital environments.
• Focus on multi-step reinforcement learning in multimodal, digital environments.

Skills

Developing and training ML models

Deep learning experience

NLP / Generative AI

Experience with reinforcement learning

Education

PhD in Computer Science or AI

Tools

Python

TensorFlow

PyTorch

JAX

Job description

Google DeepMind is hiring a Post-training Agentic Research Scientist to advance Gemini and deploy agents at scale. You will identify model improvements, train and evaluate agents across diverse digital environments, and focus on multi-step reinforcement learning in multimodal settings to accelerate real-world deployments.

You will contribute to the broader research community through publication and collaboration, joining a mission to push AI capabilities while ensuring safe and ethical