went through the charles schwab senior swe (they don't use the L5 label internally but the band roughly maps there) interview loop in spring 2026 for a position in their trading infrastructure group. system design was the most interesting part so let me break it down.
the round was 60 minutes with two engineers, one of whom was a principal. they gave me a prompt around designing a high-availability order routing system. classic fintech domain.
what they cared about: fault tolerance. they wanted explicit discussion of what happens when a downstream service is unavailable mid-transaction. partial failures, rollback, idempotency keys. this came up immediately. latency vs correctness trade-offs. in fintech you often can't sacrifice correctness for speed the way you might in, say, a social app. they pushed on this directly. audit trail. i brought up event sourcing early and they lit up. having a durable log of every state change, not just the current state, is a pattern they clearly value. regulatory compliance at the design level. not deep legal stuff but "how would you design this so that you could prove to an auditor what happened and when."
i asked a lot of clarifying questions upfront (read consistency requirements, SLA targets, team ownership model) and they seemed to appreciate that.
what i'd skip: spending a ton of time on generic horizontal scaling strategies. they already assume you know how to shard a database. they want to know if you understand the domain-specific constraints.
one thing that surprised me: they asked how i'd handle schema evolution without downtime. came up naturally from the audit log discussion. have a thoughtful answer for that.
total round: about 50 minutes of actual discussion, 10 minutes for my questions. they were generous with the Q&A time which felt like a good sign.