Turn job requirements into a structured interview scorecard
Translate duty-linked requirements into consistent evidence criteria and anchored interview ratings
10 min setupTested with:ChatGPTReviewed: 2026-08-28
1
Add your context
Your text stays in this browser. AILesson Prompts does not send it to a model or server.
2
Your prompt
Unfilled fields remain visible as placeholders, so you can still copy and edit the prompt
Build a structured interview scorecard strictly from approved, duty-linked role requirements.
Duties, outcomes, essential and trainable qualifications, decision rights, level, conditions, and exclusions:
[role]
Stages, interviewers, duration, accommodations, evidence, work samples, scale, and hiring decision:
[process]
Jurisdiction, prohibited topics, privacy, retention, weighting, vetoes, independent scoring, calibration, and approvers:
[governance]
Map each essential requirement to observable job evidence and remove duplicates. Exclude personality impressions, culture fit, pedigree, protected characteristics, and requirements without an approved duty basis. For each criterion define why it matters, evidence sources, one consistent primary question or work-sample task, neutral probes, behavioral anchors for every rating, insufficient-evidence handling, and what must not affect the score. Keep trainable skills separate from entry requirements. Set weights and vetoes only when supplied and verify arithmetic; otherwise leave them for approval. Require interviewers to capture concise evidence and score independently before discussion. Add accommodation-equivalent evidence routes without lowering the criterion. Provide calibration cases, conflict-resolution rules, an overall decision method, and an audit for overlap, adverse proxy risk, interviewer discretion, privacy, and legal/HR review.
Private by defaultPrompt assembly happens locally in your browser. Avoid placing confidential information into any AI service unless your organization allows it.
From input to outcome
A worked example
See how concrete context turns this recipe into a usable result
Actual input
Approved role requirements
Operations Coordinator. Essential: organize 20-30 concurrent requests, clear written communication, basic spreadsheet use, and evidence of improving a recurring process. Must confirm owners and acceptance criteria, maintain Linear, coordinate without changing priority, and protect routine operational data. Linear is trainable. No degree or years requirement. No people management or spend authority.
Interview process
Two stages: 45-minute structured interview with Hiring Manager and Support Lead; then 35-minute work sample reviewed by Operations Director. Rating 1-4 plus Not enough evidence. Same core questions for all candidates. Candidates may answer orally, in writing, or use assistive technology with advance arrangement. Final decision made by panel after independent scores.
Fairness and governance
UK. HR supplies prohibited-topic guidance separately. Do not collect family, health, age, nationality beyond work-right process, or other unrelated personal data. Weight proposal: request organization 30%, written communication 20%, process improvement 30%, data judgment 20%. No approved veto. Notes retained under existing HR policy; exact period not supplied. HR approves scorecard and runs calibration. Interviewers score independently before panel.
Example output
Scorecard total: 100% — request organization 30, written communication 20, process improvement 30, data judgment 20. The arithmetic is correct, but weights remain a proposal until HR approval. No veto criterion. Linear experience is not scored because it is trainable.
Request organization — 30%
Primary question: “Tell us about a time you coordinated several incoming requests with competing dependencies. How did you make ownership, acceptance, and status visible without taking over priority decisions?” Probes: volume; system used; one conflict; evidence of result.
Anchor 1: gives no clear method or confuses coordination with setting others' priorities. 2: describes a basic list but ownership, acceptance, or dependency handling is incomplete. 3: gives a concrete system with owners, acceptance, status, and appropriate escalation. 4: also demonstrates adaptation under changing constraints and evaluates whether the system worked.
Written communication — 20%
Work sample: summarize a fictional five-request handoff packet for three named stakeholders in 20 minutes, with an equivalent oral-plus-dictation route if arranged. Score clarity, factual fidelity, explicit owner/unknown, and actionable structure—not accent, typing speed, visual polish, or assistive method.
Process improvement — 30%
Question: “Walk us through a recurring process you improved: starting evidence, change, involvement, risk controls, and how you evaluated it.” Level 3 requires a specific before-state, reasoned change, stakeholder involvement, and observable follow-up; level 4 adds trade-offs and revision after evidence.
Data judgment — 20%
Scenario: a request sheet contains a customer email, an owner, missing acceptance evidence, and a proposed priority change. Candidate identifies unnecessary exposure, preserves priority authority, requests missing evidence, and escalates appropriately.
Use Not enough evidence rather than guessing. Record a short evidence locator and score independently. Calibration: compare one fluent but generic response with one less polished, evidence-rich response; the latter should score higher. Panel resolves differences by returning to anchors, not averaging away unsupported ratings. HR must approve anchors, work-sample data, accommodation process, prohibited-topic guidance, and note-retention implementation.
Why this works
1
Behavioral anchors reduce the influence of vague similarity and confidence impressions.
2
Independent evidence scoring before discussion reduces group anchoring.
Check the result
Does every scored criterion map to an approved duty and observable evidence?
Are trainable skills, accommodation routes, and insufficient evidence handled fairly?
Are weights, vetoes, decision rules, privacy, and legal approval explicit?
ChatGPT