Stryker · Primly Community

Stryker senior system design interview, what to expect: my notes from the onsite

infra_ines · 5 replies

Going through Stryker's loop for a senior SWE role (they call it "Senior Software Engineer" but internally it maps to something like a tech lead / IC5 band). The system design round was the most distinct part so writing this specifically.

The prompt I got: design a platform for tracking surgical instrument usage across hospital systems. Think: which instruments went into which OR, usage counts for wear-and-replacement scheduling, traceability for recalls.

This is very different from the classic "design Twitter" or "design a URL shortener" type of prompt. A few things that matter here that wouldn't matter in a typical SaaS system design:

Regulatory traceability. Every instrument needs a full audit trail. Not just "log it" but immutable, tamper-evident records. I talked through event sourcing as a pattern and they responded well to it.

Unique identifier schemes. Medical devices have UDI (Unique Device Identifier) standards. I didn't know this going in but had heard of it, which helped. You don't need to be an expert, but knowing the concept landed well.

Offline-first edge cases. Hospitals have OR environments that can't rely on consistent internet connectivity. The interviewer specifically probed how data syncs when an edge device comes back online.

Alert latency vs. throughput tradeoffs. Recall notification pathways are latency-sensitive; usage analytics are batch-friendly. I separated those in my design.

The interviewer wasn't trying to catch me out on medtech specifics I couldn't know. He was watching how I structured the problem and how I asked clarifying questions. I spent 10 minutes on requirements gathering before drawing anything.

One thing I'd do differently: I rushed into the data model too fast. Should have spent more time on the failure modes first. The interviewer circled back to ask about failure modes anyway.

For Stryker senior system design, the core skill is designing around reliability, traceability, and constraints, not just throughput. Know your CAP theorem basics, know event sourcing, and be comfortable with audit-log patterns.

5 replies

ux_uma

the offline-first + sync-on-reconnect probe is a great question honestly. I design data pipelines for a living and even I'd have to think carefully about idempotency at the edge there. did they care what specific tech you used for the sync layer or was it conceptual?

sec_sasha

totally conceptual. I mentioned change data capture and event queues without naming a specific vendor and that was fine. they weren't looking for "Kafka vs. Kinesis" they were looking for whether you understood the problem.

sdr_sky

honest question: how much of this is actually on the job vs. just interview theater? do Stryker SWEs really think about UDI standards day to day, or is that more of a compliance team concern?

marketer_mei

from what I learned during my loop, the software team works closely with quality/regulatory but doesn't own those standards directly. you'd need to know enough to build the right data model and respect the requirements. not be an FDA expert yourself.

quietquit_quincy

saving this. interviewing at a medtech adjacent startup next month and this framing around traceability and audit trails is exactly what I needed to think through before that system design round.