Just finished a senior SWE loop at Sony (SIE, San Mateo) in early 2026. Wanted to write this up while it's fresh because there wasn't much out there when I was prepping.
The system design round was 60 minutes with two interviewers. One was the hiring manager, one was a staff engineer from a different team. Format: About 5 minutes of small talk and role background The problem: design a real-time notification delivery system at scale. They said tens of millions of PlayStation Network users, mix of push, email, and in-app Remaining time: whiteboard on Miro (remote interview), me driving, them asking questions throughout
What they actually cared about:
Reliability over throughput. When I started optimizing for latency, the staff eng redirected me. PSN has outages that make headlines. They want to know you've thought about graceful degradation, retry logic, circuit breakers.
Data modeling. They drilled into my schema choices longer than I expected. What's in Postgres vs Redis vs Kafka. Why.
Cross-team dependencies. At one point the HM asked how my design would handle a downstream team's API changing without notice. Good signal this is a real pain point for them.
They didn't ask me to estimate QPS in the first 2 minutes like FAANG companies often do. More exploratory, fewer numbers on the whiteboard.
I was leveling for what I'd call L5-equivalent (they use their own internal levels, not the L1-L9 Google ladder). If you're interviewing at the senior level, expect the system design to be a real conversation, not a performance.
One thing I wish I'd prepped more: failure modes. How do you handle notification storms when a big game drops. How do you prioritize. Have an answer for that.