Apply AI/ML to fault-tolerance and reliability across Google's global data center fleet

Likely format

Google SWE loop: phone screen(s) covering coding and ML/infra domain knowledge, followed by onsite rounds spanning coding, ML system design, and Googleyness/leadership behavioral interviews.

What this interview tests

• ML model design for fault prediction/anomaly detection at fleet scale
• Working with large-scale telemetry and operational data pipelines
• ML infrastructure: model deployment, evaluation, fine-tuning, debugging
• Cross-functional partnership with hardware designers and SREs
• Technical leadership: setting direction, translating findings for executives

Common question themes