PhD engineer building end-to-end ML systems across Google's stack

This is a 2026-start, early-career PhD role open across multiple Google product areas (AI & Infrastructure, Cloud, YouTube, Search, Ads). The work spans the full ML stack — from low-level hardware acceleration and compiler optimizations up to model architecture and production APIs — plus optimizing production system performance (bottlenecks, memory inefficiencies, errors) and writing well-tested, reviewed code. Preferred experience includes deep learning frameworks (TensorFlow/JAX/PyTorch), model architectures (CNNs, NLP transformers, diffusion/vision transformers), and building a full AI application stack from data pipelines to user-facing APIs.

Step into this interview

Free · a live voice mock calibrated to this exact role

Practice this interview

Likely format

Google standard loop: phone screen(s) with coding, followed by onsite rounds covering coding/algorithms, ML/systems design, and Googleyness/leadership behavioral interviews

What this interview tests

End-to-end ML system design: data pipeline to model to production API
Deep learning frameworks and model architecture tradeoffs (CNNs, transformers, diffusion)
Diagnosing and fixing performance bottlenecks/memory inefficiencies in production ML systems
Research-to-production translation of ML/AI work at large scale
Coding fundamentals in Python/C/C++/Java/Go with well-tested, reviewed code