Back to problems

Existing Codebase Debugging Round

AI Coding · Scale AI · Medium

Requirements The candidate is handed a small Python codebase that already runs end to end: four source files plus a demo entry point, with a test suite that asserts the intended behaviour of every public method. The workspace models an annotation batch pipeline: Annotator, Task / TaskPool, Scorer, and Batch. A main.py exercises the flow against three annotators in two courses and five tasks. Everything is standard library; nothing to install. Four defects have been seeded…

Checking your access…