A system design round, two reports tagged Cursor (November 2025 and April 2026). The April one comes from a stacked pull-request review startup acquired by Cursor with its staff in New York, which matches Graphite; the November one from a pull-request tool startup's phone screen. In the April report: design internationalisation for a large website, Airbnb-like, with 10 million monthly active users and 200 engineers, going from English only to many languages. Optimise latency, and keep the change for engineers as small as possible, for example wrapping text as T("Hello"). Asked in a reply, the candidate added it is a regular design that has to consider what developers and translators need. In the November report the scope was small: how a web page does i18n when different words have to be replaced, how to store them and how to fetch them.
Cover happy path. Clarify scope. Identify the obvious bottleneck. Pick a reasonable storage and reasonable scaling approach.
All of the above plus: explicit failure handling, durability vs latency trade-offs, choose the right batching/caching strategy, articulate why.
All of the above plus: organizational concerns (rollout, migration, on-call), quantitative analysis, multi-region considerations, what could go wrong with the proposed solution at 10x scale.
Interviewer hints: Optimise latency, and keep the change for engineers small, e.g. T("Hello") (the brief, one report).; Consider what developers and translators each need (the candidate, in a reply).