Code RoomSurvey response analyzer
MediumPrep Room Coding #50

Survey response analyzer

Vibe & agenticAI agents & automationMid–Senior~14 min

Your quarterly developer-experience survey collects hundreds of free-text answers, and leadership wants themes. Design an agent that analyzes it. Those answers were written under a promise of anonymity, and some name names or describe situations only one person could have lived — so what guardrails does the agent enforce before any output reaches a manager? How do you verify the themes are faithful and nothing traceable leaks, and how do you pitch this so engineers still trust the survey next quarter?

Implement
screen_survey_themes(theme_names: list[str], respondent_counts: list[int], has_identifiable_detail: list[bool], k_floor: int) → list[str]
Examples
in[["flaky ci","laptop specs","review turnaround"],[12,7,3],[false,false,false],5]out["flaky ci:release","laptop specs:release","review turnaround:suppress"]
in[["oncall load","manager named in quote"],[9,6],[false,true],5]out["oncall load:release","manager named in quote:paraphrase"]
in[[],[],[],5]out[]
What a strong answer looks like

Turn the fuzzy goal into a bounded agent task. Say what the agent reads, what it may do on its own versus draft for a human, and what it must never touch. Then earn the trust: how you’d verify it before rollout, and the pitch (the problem, what you built, why it matters).

0:00 of about 14 min

Agent build: design the agent in plain language (or narrate it). The coach grades the decomposition, the guardrails, the verification plan, and the pitch.

Which questions mattered is sealed until you submit. Telling you now would just be handing over the edge cases.