1000 users doing what?
Reading a blog? Joining a livestream? Placing bids? Moving money? Uploading video? Calling 911?
The candidate is wrong for jumping straight to Kubernetes, Kafka, Redis, sharding, and multi-region.
But the tweet is wrong too.
It mocks overengineering while ignoring the actual mistake: nobody defined the use cases, constraints, traffic shape, or SLOs.
That is the real system design failure.
Interviewer: Do you know System Design?
Candidate: Yes, absolutely!
Interviewer: Okay, design a system for 1000 users.
Candidate: Kubernetes, microservices, Kafka, Redis, sharded MongoDB, CDN, multi-region deployment...
Interviewer: stares
Interviewer: I said 1000 users. Not 100 million.
Candidate: Yeah but... what if it goes viral?