Software Engineer, PhD, Early Career, AI/Machine Learning, 2026 Start: Google

This early career role is specifically designed for PhD graduates, leveraging deep research expertise to develop the next generation of AI and Machine Learning technologies at a massive scale. You will work on a specific project critical to Google’s needs, spanning various product areas like AI & Infrastructure, Cloud, YouTube, Search, and Ads. Google is a leading producer and consumer of ML/AI technology, offering opportunities to work with custom ML hardware infrastructure.

Responsibilities
System Development: Collaborate or lead on projects to carry out design, analysis, and development of advanced ML systems across the stack using your research expertise.

Full-Stack Implementation: Support building end-to-end ML Systems that involves working across the full stack, from low-level hardware acceleration and compiler optimizations to high-level model architecture and production APIs.

Performance Optimization: Optimize complex system performance by analyzing and fixing bottlenecks, memory inefficiencies, and errors in production systems to meet stringent customer goals.

Engineering Excellence: Elevate engineering excellence by writing well-tested code, conducting code reviews, and fostering a culture of quality.

Qualifications
Minimum Qualifications
Education: PhD degree in Computer Science, ML/AI, or a related field, or equivalent practical experience.

Coding: Experience coding in one of the following programming languages including but not limited to: Python, C, C++, Java, JavaScript or Golang.

Core Knowledge: Experience in Machine Learning or Artificial Intelligence.

Preferred Qualifications
Research Focus: Research experience in designing, developing, or applying ML/AI systems or applications in a large-scale distributed environment.

Modeling: Experience in designing, training, or refining complex ML/AI models.

Frameworks: Experience in deep learning frameworks like TensorFlow/Jax/PyTorch.

Full Stack: Experience in building a stack for an AI-powered application, including data ingestion and processing pipelines, building APIs, and connecting the model to a user-facing interface.

Architecture: Familiarity with model architectures (e.g., CNNs, NLP Transformers