Design a text-to-video service like sora
PremiumLLM systemHard
- Text to video
- GPU job scheduling
- Weighted fair queuing
Quick answer
This LLM system interview question asks you to design a text-to-video service like sora. Aim for a design that uses language models to serve users well, then explain retrieval, prompting, serving, evaluation and trade-offs. Interviewers want to see how you control quality, cost and latency, and how you evaluate the results.
The problem
Design a text-to-video service like Sora: users type a prompt and get back a short generated video. Each job holds one GPU for minutes, so the hard parts are fair queueing across users and tiers, progress and cancellation, safe retries, safety checks and cost.
Users write a prompt, may attach a starting image, pick a length (5, 10 or 20 seconds) and a resolution (480p or 720p), and later watch, download, share or delete the result. Evening demand is twice the daily average while the GPU fleet is fixed, so queues form every day.
What must a text-to-video service like sora do?
- Submit and track: Users submit a prompt and see their queue position, then progress by step, then the finished video.
- Fair queueing: Jobs are scheduled fairly across users and tiers; no single user or tier can take the whole fleet.
How well must it do it?
- Queue wait: Pro jobs start within 2 minutes and Plus jobs within 10 minutes at p95, even at peak. Free jobs show an estimate and may wait hours at peak.
- No wasted GPU time: A blocked prompt never reaches a GPU, and a failure loses at most 10 steps of work.
What scale should you design for?
Use these numbers in your estimates.
The scale numbers unlock with this questionUnlock
Key concepts
- Text to video
- GPU job scheduling
- Weighted fair queuing
- Diffusion checkpointing
- Content safety
- C2PA provenance
How you’ll practise
- Problem framing
- Data and retrieval
- Prompting and orchestration
- Evaluation and safety
- Serving, cost and latency
- Wrap-up
Lala, personalized to you, gives a hint when you ask and reviews your design when you are ready.
What do interviewers look for?
In any LLM system design round.
- Defines the user task and what a good answer is
- Explains how relevant context reaches the model
- Plans evaluation before and after launch
- Controls cost and latency with caching and routing
- Covers safety, privacy and abuse
- Handles failures and fallbacks gracefully
Questions people ask about this interview question
- Which design type is this question?
- It is an LLM system question. You design a text-to-video service like sora and explain your choices the way you would in an interview.
- Is this question asked at entry level?
- It is marked Hard. It suits engineers preparing for senior interviews, where you are expected to drive the discussion and go deep on trade-offs.
- What should I cover first?
- Start with the user task, what a good answer looks like and how it will be measured. Agree on scope with the interviewer before you design anything.
- Can I practise it for free?
- It is a premium question. Sign up and unlock it once for 10 credits; it then stays unlocked. When you finish, Lala, your personalized AI mentor, can review your design for 30 credits.
Related questions
Updated · Written by the SevyDevy team