All interviews
Google logo

Google

Senior

Google Senior SWE — build eye-tracking perception software for Android XR devices

This role is on Google's Android XR / XR Perception team, writing compute-constrained perception software (eye tracking specifically called out) that runs at high frame rate and bandwidth on mobile XR devices. The bar is senior: 5+ years of software development, 3+ years shipping/maintaining production software, 1+ year of design/architecture, plus required computer vision/AR experience, with eye-tracking-on-XR-devices and camera sensor pipeline experience called out as strongly preferred. Expect deep technical discussion on embedded/on-device ML perception systems, performance-constrained software design, and cross-team collaboration, alongside Google's standard coding and system design loop.

Step into this interview

Free · a live voice mock calibrated to this exact role

Practice this interview

Likely format

Standard Google SWE loop: recruiter screen, phone screen(s) with coding, then onsite virtual loop of 4-5 rounds covering coding/algorithms, system design, a domain-specific (perception/CV) technical round, and Googleyness/leadership, followed by Hiring Committee review.

What this interview tests

  • Perception software for compute-constrained, high-frame-rate Android devices
  • Computer vision / AR / eye-tracking domain experience
  • Camera sensor pipelines and data flow on Android devices
  • Architecture/design documentation and cross-team technical communication
  • Code review practices and engineering process improvement
  • Collaboration with ML engineers to productionize perception models on-device

Common question themes

Design a perception pipeline that must hit a real-time frame budget on constrained hardware

Describe hands-on experience with eye tracking, gaze, or camera sensor pipelines

Tell me about a design doc you wrote that shaped how your team built a system

How do you approach code review to improve team practices, not just catch bugs

Describe collaborating with ML/research engineers to get a model running efficiently on-device

A time you diagnosed and fixed a performance bottleneck in resource-constrained software

View the original posting

Related interviews