Your question is Scale Audio Delivery Infrastructure. Take a moment with it on the right.
Talk me through your thinking if you like. When you're confident, submit your answer and I'll grade it like a real screen (7/10 or better passes).
You are the engineering manager responsible for scaling the media delivery layer behind a large audio platform. The system must support millions of daily downloads and streams across mobile, web, and browser extension surfaces, including generated narration files and premium content. Traffic is highly spiky around commuting hours and new feature launches, and the current setup has started to show elevated latency, cache misses, and rising egress costs. At the same time, product leadership wants faster global playback start times, finance is pushing for tighter infrastructure efficiency, and support has escalated complaints about intermittent failures on poor networks. You need a design that can scale reliably without overbuilding, while preserving room for phased rollout and operational recovery if the migration causes regressions.
| Detail | Value |
|---|---|
| Daily audio requests | 18 million |
| Peak read traffic | 220K requests/minute |
| Audio catalog size | 95 million files |
| Surfaces supported | iOS, Android, Web, Chrome Extension |
| Playback start SLA | p95 under 250 ms for cached assets |
| Availability target | 99.95% |
| Infra budget increase cap | 15% QoQ |
| Delivery deadline for phase 1 | 12 weeks |
How would you design and execute a high-throughput media delivery network that handles millions of audio file downloads and streams daily?