Model tier for batch classification
You're building a Python pipeline that classifies 2 million support tickets into one of 12 routing buckets, then for the ~3% flagged 'escalation' it drafts a nuanced customer-facing apology. A teammate proposes calling the single largest/most-capable model for every ticket to 'keep it simple.' What's your read on the model-tier choice, and how would you structure it?
Implement
route_ticket_tiers(tickets: list[str], confidence_floor: int, premium_budget: int) → list[str]Examples
in
[["billing|95|0","refund|88|0","billing|91|1"],70,2]out["cheap","cheap","premium"]in
[["billing|40|0","unknown|99|0","shipping|70|0"],70,1]out["premium","queue","cheap"]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 15 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.