HR teams increasingly use AI to work faster: screening CVs, sharpening job descriptions, creating interview questions, and structuring onboarding.
The difference between "AI that produces some text" and real time savings lies in prompt engineering: clear instructions, defined output formats, and good review steps.
TipPrivacy tip: do not enter personal data (anonymise CVs and cases) and only use approved tools within your organisation's policy.
Template 1 — CV screening (role-based)
- Goal: quickly and consistently assess against job requirements, plus get targeted interview questions.
- Prompt: "You are an experienced HR analyst. Here are (anonymised) job requirements: [paste requirements]. Here is an (anonymised) CV summary: [paste CV summary]. Assess the match and provide: (1) score 0–100, (2) top 5 strengths vs requirements, (3) top 5 gaps/risks, (4) 5 targeted interview questions to test the gaps. Answer in a table."
Template 2 — Improve a job posting (more inclusive and sharper)
- Goal: make job postings more attractive without vague buzzwords.
- Prompt: "You are a recruitment copywriter. Rewrite this job posting: [paste text]. Goals: (1) more concrete about tasks and success criteria, (2) inclusive language, (3) less jargon, (4) clear 'must haves' vs 'nice to haves'. Deliver: (a) improved version, (b) list of 10 reasons why this is better, (c) 5 alternative job titles."
Template 3 — Interview questions based on competencies
- Goal: conduct better interviews with behavioural questions and a scorecard.
- Prompt: "You are an HR business partner. For this role: [role]. Competencies: [list]. Create 12 interview questions using the STAR method, distributed across the competencies. For each question provide: what a strong vs weak answer looks like. Also create a scorecard (1–5) with assessment criteria."
Template 4 — Onboarding plan (30/60/90 days)
- Goal: consistent onboarding with measurable objectives.
- Prompt: "You are an onboarding coach. Create a 30/60/90-day plan for: [role]. Context: [team/goals]. For each period provide: (1) objectives, (2) key tasks, (3) stakeholders to meet, (4) deliverables, (5) success metrics. Keep it practical and concrete."
Template 5 — Performance review preparation
- Goal: make reviews more objective and concrete.
- Prompt: "You are an HR advisor. Use this input: goals: [goals], results: [results], feedback: [feedback], challenges: [challenges]. Create a draft review with: (1) summary, (2) strengths with examples, (3) areas for improvement with examples, (4) development plan (3 actions), (5) proposed goals for the coming period. Write in a neutral, factual tone."
Tips: how HR gets the most out of AI (without risk)
- Anonymise input: no names, addresses, dates of birth, or unique identifiers.
- Ask for structure: tables, scorecards, and explicit criteria.
- Ensure bias checks: have AI explicitly flag potential biases.
- Use AI as a "first draft", not as the final decision-maker.
- Document what is and isn't allowed in your AI policy (prevent Shadow AI).
Conclusion: make AI quality repeatable
With standardised templates, AI usage becomes predictable: faster, more consistent, and easier to review.
Want to embed this organisation-wide? Qrio helps with training, assessment, and demonstrability — so productivity and compliance go hand in hand.