Research Scientist, Gemini Architecture, Generative AI

Google•Mountain View, CA
•$147,000 - $210,000

About The Position

Google’s research organization pursues fundamental research, product innovation, product contributions, and infrastructure goals. Research Scientists set up large-scale tests, quickly and broadly deploy promising ideas, manage deadlines and deliverables, and apply current theories to develop products, processes, and technologies. The work can include creating experiments, prototyping implementations, and designing architectures to address problems across computer science. Research Scientists contribute to the wider research community by sharing and publishing findings, drawing on internal projects and collaborations with partner universities and technical institutes. Google Research conducts fundamental research and influences product development, with teams publishing in academic journals, releasing open-source projects, and applying research to Google products. This role focuses on Transformer and Agentic LLM architectures and algorithms for current and future AI products, understanding the structure of real-world AI tasks, and using research findings to improve Google products and AI solutions.

Requirements

  • PhD in Computer Science, a related field, or equivalent practical experience.
  • Programming experience in Python.
  • Experience applying Transformer models to natural language processing or multimodal machine-learning tasks.
  • Experience developing, fine-tuning, or serving large language models (LLMs).
  • One or more scientific publication submissions to conferences, journals, or public repositories, such as CVPR, ICCV, NeurIPS, ICML, or ICLR.

Nice To Haves

  • Two years of coding experience.
  • One year of experience owning and initiating research agendas.

Responsibilities

  • Understand Transformer and Agentic LLM architectures and algorithms conceptually, theoretically, and analytically, and develop improvements for current and future AI products.
  • Understand the structure of real-world AI tasks, potentially formalize that structure, and create similar datasets to improve AI solutions.
  • Impact Google products based on research findings.
  • Carry out and publish fundamental scientific advancements.
  • Be familiar with Transformer code, Python, JAX, LLM training, reinforcement-learning fine-tuning and algorithms, and evaluation measurements.

Benefits

  • Equity
  • Benefits
© 2026 Teal Labs, Inc
Privacy PolicyTerms of Service