Cutting Realtime Match Reads in the Queue
How an expensive lobby fetch turned two players queuing into thousands of reads, and how moving selection back to the server fixed the cost shape.
Evidence-backed writeups on the systems behind AlgoArena: judging, telemetry, assessment validation, metrics, and the product loops that are stable enough to explain publicly.
A storage-first cost model for replayable coding matches, written before keystroke playback became part of the product surface.
Why exact-output judging breaks on permutation tasks, and how AlgoArena started accepting structurally valid answers instead.
How UI assessment review moved from code-only inspection toward DOM snapshots, viewport coverage, console issues, and visible artifacts.
A small analytics methodology for counting real product activity across practice, puzzle rush, solved problems, assessments, and classroom sessions.
Technical evidence and systems notes from completed work.