Google's AI coding agent, Jules, aims to transform developers from solo contributors into orchestrators of intelligent systems. The Labs team focuses on incubating early-stage efforts to discover and create new ways to advance Google's core products through exploration and the application of new technologies, building new solutions that have the potential to transform how users interact with Google and deliver significant impact over longer timeframes.
Requirements
- 2 years of experience with experience with machine learning algorithms and tools (e.g., TensorFlow), artificial intelligence, deep learning, or natural language processing.
- 1 year of experience with software design and architecture.
- 5 years of experience with data structures/algorithms.
- Experience developing, launching products, or technologies with Large Language Models (LLMs).
- Experience developing accessible technologies.
Responsibilities
- Design and improve core agent cognitive architecture for Jules, including planning, execution, instruction following, feedback, memory, state management and tool-use capabilities.
- Improve agent state structure and update, using context engineering, and agent self-correction loops, using code verification and feedback to ensure high quality and correctness.
- Lead efforts to benchmark performance on Gemini and Jules benchmarks, towards metrics and capabilities.
- Own quality and reliability experience aspects related to agent performance, compliance and consistency.
- Collaborate closely with product, infrastructure, and research teams to ship new agent capabilities and deliver a reliable product experience.
Other
- Bachelor’s degree or equivalent practical experience.
- 2 years of experience in testing, maintaining, or launching software products.
- Master's degree or PhD in Computer Science or a related technical field.
- 1 year of experience in a technical leadership role.
- Google is proud to be an equal opportunity workplace and is an affirmative action employer.
