ETL pipeline decomposition
You need a nightly ETL pipeline that pulls events from an S3 export, dedupes and enriches them, and loads a Snowflake analytics table other teams depend on. How would you break this into agent-delegatable pieces? Which parts do you delegate, which do you own, and where do the human checkpoints sit?
Implement
dedupe_events(event_lines: list[str]) → list[str]Examples
in
[["e2|2026-01-02T00:00:00Z|b","e1|2026-01-01T00:00:00Z|a"]]out["e1|2026-01-01T00:00:00Z|a","e2|2026-01-02T00:00:00Z|b"]in
[["e1|2026-01-01T00:00:00Z|old","e1|2026-01-03T00:00:00Z|new","e1|2026-01-02T00:00:00Z|mid"]]out["e1|2026-01-03T00:00:00Z|new"]What a strong answer looks like
Treat the AI’s output as a draft to verify, not an answer to trust. Name the specific flaw and the input that triggers it, say how you’d catch it (tests, edge cases, reading critically), and how you’d re-prompt or decompose to get it right.
0:00 of about 24 min
Vibe & agentic: describe the solution in plain language (or narrate it) and the coach grades your approach.
Which questions mattered is sealed until you submit. Telling you now would just be handing over the edge cases.
Run or narrate your approach, then ask the coach.