Length 45 minFormat Video callThey grade Past projects and the reasons behind your technical decisions
Length 60 minFormat Live coding, one or two sessionsThey grade A medium-to-hard problem solved quickly, with tradeoffs and complexityHeads-up The screens filter hard on speed.
Length variesFormat Live coding, usually two roundsThey grade Correctness and code quality; test before you call it doneHeads-up Harder than the screen. Reported topics: graph traversal, a Connect Four winner check, event stream processing.
Length 120 minFormat Build a small working module from an empty fileThey grade Running, well-organized, extensible codeHeads-up Past prompts: a table module with joins, a stream processor.
Length 60 minFormat Design discussionThey grade Search systems at depth: indexing many workplace apps, permission-aware results, multi-tenancy, ranking
Length 60 minFormat ConversationThey grade Ownership, conflict, influence and what you personally changedHeads-up Some candidates also report an exercise that watches how you use AI coding tools on ordinary work.
