YOUR BRIEF
What to work through
Work through a text-to-video service as a spoken system design interview. Start by clarifying the scope, sketch a working design and explain the trade-offs behind your choices.
- Submit a prompt with settings such as duration and resolution, and get a job id back
- See a job’s status and an honest estimate of the wait
- Cancel a job that is queued or running
- Watch and download the finished clip
BEFORE YOU START
Constraints to discuss
- Each job runs on exactly one GPU for minutes, never split across GPUs
- Fair share: no single user can take the whole fleet
- Bursts far above GPU capacity queue up instead of failing
- A worker that dies mid-job does not send the job back to zero
Need a starting point?
Watch for these common gaps as you prepare:
- A synchronous request held open for minutes
- Job state only in the scheduler’s memory
- Counting fairness in jobs instead of GPU time
- No fencing, so a slow worker overwrites a retried job’s result
- Cancelling only queued jobs, leaving running ones on the GPU
Why this question is in the collection
These sources connect the problem to company interview reports or preparation guides. They do not guarantee what a company will ask you.