30 AI Prompts for HR & People Teams
Copy-ready prompts that assume you will attach the CV, the policy or the review notes — built to produce fair, structured, on-brand people work, not bland filler.
Each prompt gives the AI an HR persona, tells it what you will attach, and specifies the framework and format so the draft is usable and defensible. These prompts are tool-agnostic and optimised to work across all major AI assistants — Claude, Microsoft Copilot and ChatGPT (and Google Gemini). Replace the [bracketed placeholders] with your own details, and always review AI output before you act on it.
Recruitment and job descriptions
Inclusive Job Description
Posting a new role and need it to convert, not just list duties
no file - role, industry, location details
- Write a one-line hook and explain why the role matters
- List 12-month outcomes owned, not a task list
- Split must-have vs nice-to-have requirements, add an honest team and pay note
- Use gender-neutral language, Malaysia context, no unnecessary degree requirements
Full job advert plus a 120-word LinkedIn version
See example AI output
Hook: Build the finance stack a fast-growing F&B chain can actually trust. Why it matters: You will own reporting the CEO reads every Monday. 12-month outcomes: Close the books in 3 days, launch a new forecasting model. Must-have: 5+ years FP&A. Nice-to-have: Power BI. Team and pay: Reports to CFO, small 4-person team, RM7,000-9,500. LinkedIn version (120 words): Shorter hook-led summary with a link to apply.
Sourcing Plan and Boolean Search
Building a sourcing plan for a hard-to-fill role
no file - role title, must-have skills, location
- Rank ten sourcing channels by likely yield
- Write a ready-to-use LinkedIn boolean search string
- Name three niche communities and one outreach hook for passive candidates
- Keep it realistic for an in-house recruiter, no paid tools assumed
Ranked channel table plus the boolean string and outreach hook
See example AI output
Top channels: 1) LinkedIn, 2) JobStreet, 3) Ricebowl developer community, 4) alumni networks, 5) referrals. Boolean: (backend engineer OR software engineer) AND (Node.js OR Python) AND (Malaysia OR KL) -intern Communities: KLDevs Slack, Malaysia Python User Group, Girls in Tech KL Outreach hook: Saw your work on [project], we are solving a similar scaling problem at [company], worth a 15-minute chat?
Job Ad Rewrite for Employer Brand
Turning a generic job ad into one that reflects your employer brand
current job description draft
- Sharpen the opening hook and convert listed duties into outcomes
- Cut clichés and generic phrasing
- Add one specific, honest reason to join
- Keep every factual requirement and flag anything that reads as a red flag to candidates
Rewritten advert plus a two-bullet note on what changed and why
See example AI output
Rewritten hook: Own the checkout experience for 2 million monthly shoppers. Outcomes: Ship 3 major checkout features, cut cart abandonment by 5 percent this year. What changed: - Replaced the fast-paced-environment cliché with the real 2M-shopper scale, more credible to senior candidates - Flagged: original ad listed 15 responsibilities with no seniority signal, may deter senior applicants, recommend trimming to top 6
Defensible Pay-Range Rationale
Justifying a proposed salary band before it goes to the hiring manager
role, level, location, market salary data
- List the factors that should set the pay band
- Recommend how to position against market and suggest non-cash levers
- State assumptions clearly and flag confirming with local benchmark data
Suggested band structure plus a short justification memo
See example AI output
Suggested band: RM6,500-8,200 for a Senior Analyst in Kuala Lumpur. Factors: internal equity with Analyst II, assumed market scarcity, 12 percent premium for niche SQL and Python skills. Non-cash levers: hybrid work, learning stipend, extra annual leave. Memo: Band set at market median plus 5 percent given a tight talent pool; confirm against current benchmark data before finalising. Assumes a standard 13-month salary structure.
CV Screening Against Scorecard
Triaging a large batch of CVs against a role scorecard
batch of CVs, role scorecard
- Screen each CV against the must-have criteria only
- Rate fit as high, medium or low with evidence
- Write two probing questions per shortlisted candidate
- Flag bias-prone inferences and never auto-reject
Ranked shortlist table (Candidate, Fit, Evidence, Questions) plus who to interview first
See example AI output
Candidate: Tan W.L. Fit: High. Evidence: 6 years SAP FICO, led 2 migrations. Questions: How did you handle stakeholder pushback, what was the rollback plan? Candidate: Nurul A. Fit: Medium. Evidence: 3 years, no migration lead experience. Questions: Have you scoped a migration solo? Interview first: Tan W.L., strongest evidence match, no bias flags found in the notes.
Interviews and screening
Structured Interview Kit (STAR)
Building an interview scorecard before a hiring loop starts
no file - just the role title
- Define five competencies to assess
- Write two STAR behavioural questions per competency
- Build a 1-to-5 rubric describing weak, adequate and strong answers
- Keep questions answerable by evidence, not opinion, and legally safe for Malaysia
Question bank grouped by competency plus the scoring rubric
See example AI output
Competency: Stakeholder Management Q1: Tell me about a time you managed a difficult stakeholder during a project. Q2: Describe a situation where priorities conflicted between departments. Rubric: 1-2 Weak, vague with no specific outcome. 3 Adequate, clear situation with a modest result. 4-5 Strong, specific conflict, structured resolution, measurable outcome and reflection on what was learned.
Tailored Interview Plan
Prepping a panel interview for a specific candidate, not a generic script
candidate's CV, role scorecard
- Identify specific gaps and claims in the CV to probe
- Write three questions customised to the candidate's background, not generic ones
- Flag the one risk to test hardest
Timed run-sheet with the questions and what a good answer proves
See example AI output
0-5 min: Rapport. 5-20 min: Your CV shows 8 months at Company X before leaving, walk me through that decision. Proves: honesty and self-awareness. 20-35 min: You list led migration, what was your specific role versus the team's? Proves: real ownership versus an inflated title. Biggest risk to test: the short-tenure pattern, probe retention risk directly.
Interview Notes to Scorecard
Turning messy interview notes into a defensible hire decision
raw interview notes
- List strengths and concerns, each backed by specific evidence
- Rate each competency and give a hire, no-hire or hire-with-conditions call
- Separate observation from inference and don't over-weight a single moment
Structured scorecard plus a two-sentence recommendation
See example AI output
Strengths: Strong technical depth, explained caching strategy unprompted; calm under the pressure question. Concerns: Vague on the team-conflict example, deflected twice. Ratings: Technical 4/5, Communication 3/5, Collaboration 2/5. Recommendation: Hire-with-conditions, pair with a strong lead during the first 90 days to monitor collaboration, since one weak data point should not be over-weighted.
Bias Check on Scorecard
Reviewing a completed interview scorecard before a hiring decision is finalized
draft interview scorecard
- Spot language showing affinity or halo effects
- Flag criteria not tied to the job
- Check for inconsistent standards applied between candidates
- Suggest a neutral rewrite for each flag
Flag list (Issue, Why it is a risk, Suggested rewrite)
See example AI output
Issue: Great culture fit, felt like one of us. Why it is a risk: affinity bias, not job-related. Suggested rewrite: Demonstrated alignment with the team's collaborative working style, evidenced by X. Issue: Not very polished noted for Candidate B, no such note for Candidate A. Why: inconsistent standard. Suggested rewrite: apply the same communication rubric line to every candidate.
Reference-Check Call Script
Running a reference check and want to verify specific interview concerns
role, candidate details, interview concerns to test
- Frame the call professionally
- Write six open questions on real performance and working style
- Add two questions that gently test the concerns raised at interview
Call script plus a note on what each answer should confirm
See example AI output
Framing: Thanks for your time, this is a routine reference check for [candidate] for a [role] position. Q1: How would you describe their day-to-day working style? Confirms: self-reported style matches reality. Q2: Can you share an example of how they handled a tight deadline? Concern-test Q: We noticed a short tenure at [company], what led to their move? Confirms: retention risk.
Onboarding
30-60-90 Day Onboarding Plan
Structuring a new hire's first three months around outcomes, not tasks
no file - new hire's role, reporting manager
- Define the outcome for each of the 30, 60 and 90-day phases
- List key milestones and who the new hire should meet per phase
- Write manager check-in questions
- Keep it outcome-led, not task-led, with a realistic SME ramp
Three-phase plan table plus a manager's check-in guide
See example AI output
Day 30 outcome: Understands the product and can navigate systems unaided. Milestone: complete first ticket end-to-end. Meet: team lead, QA, product owner. Manager check-in: What's still unclear about our workflow? Day 90 outcome: Owns a small feature independently. Check-in: Where do you feel you're contributing real value now?
First-Week Onboarding Schedule
Planning a new hire's literal day-by-day first week
no file - new hire's role
- Schedule IT and system access setup, plus statutory setup for EPF, SOCSO, EIS and tax
- Plan team introductions and assign one small task to complete by Friday
- Keep the pace realistic and protect focus time
Day-by-day schedule with owners
See example AI output
Mon AM: laptop, email, Slack setup with IT. Mon PM: EPF, SOCSO and tax forms with HR. Tue: meet team lead and three teammates, 30 minutes each, no more. Wed: shadow a live client call. Thu: two focus hours to start the small task, updating the FAQ doc. Fri: present the small task to the manager, the first early win.
New-Hire Policy FAQ
Translating a dense HR policy into something a new joiner will actually read
dense HR policy document
- List the 12 questions a new joiner actually asks
- Answer each in two to three sentences, in plain language
- Note where to find more detail and flag anything needing the handbook's exact wording
Titled FAQ document
See example AI output
New Joiner FAQ. Q: How many annual leave days do I get? A: 16 days a year, pro-rated from your start date. Full details in Handbook Section 4.2. Q: Can I work from home? A: Yes, up to two days a week with manager approval. Flag: exact eligibility wording should follow handbook clause 5.1 verbatim before publishing.
Warm New-Hire Welcome Email
Sending a new hire something human before their first day
no file - new hire's start date
- Cover first-day logistics, what to bring, and dress code
- Name who will greet them
- Add a friendly line that eases nerves
- Keep it warm, not corporate, and under 200 words
Ready-to-send welcome email
See example AI output
Subject: Can't wait to have you, [Name]! Hi [Name], we're counting down to [date]! Just bring yourself and your IC, we'll handle the rest. Dress is smart-casual, jeans totally fine. Sarah from the team will meet you at reception at 9am with coffee in hand. Don't worry about knowing everything on day one, nobody does. See you soon! Warmly, [Manager]
Onboarding Buddy Programme Guide
Setting up a peer buddy to support a new hire's first month
no file - generic for any assigned buddy
- Outline what to cover in weeks 1, 2 and 4
- List five proactive questions the buddy should ask the new hire
- Define signals that onboarding is going well or off track
- Keep it low effort for the buddy but high signal
One-page buddy checklist
See example AI output
Week 1: show them where things are, grab lunch together. Week 2: ask what's still confusing. Week 4: ask if they feel set up to succeed. Ask: how's the workload feeling, any tools still confusing, have you met everyone you need to. Good signal: asking questions unprompted. Off-track signal: going quiet in team chat by week three.
Policies and communication
Clear, Fair HR Policy Draft
Drafting a new HR policy from scratch, e.g. flexible work
no file - just the policy topic
- State purpose and scope, then write the rules in plain language
- Define employee and manager responsibilities
- Note exceptions and flag every point needing legal review
Policy document plus a highlighted list of items to verify with an adviser
See example AI output
Flexible Work Policy. Purpose: support work-life balance while meeting business needs. Rules: up to two remote days a week, core hours 10am to 4pm. Manager responsibility: approve requests within three days. Verify with adviser: whether remote days affect EPF and SOCSO reporting, and whether the policy needs to reference the relevant Employment Act provisions.
HR Announcement Rewrite
Tightening a rough HR announcement before it goes out to staff
rough draft of the HR announcement
- Lead with what is changing and why
- Spell out required employee actions and deadlines
- Pre-empt the top three likely questions
- Remove ambiguity and keep the tone empathetic
Rewritten announcement plus a short FAQ
See example AI output
Subject: Update to our claims process, action needed by Aug 15. Starting September, all claims move to the new portal to speed up reimbursement. Please register by Aug 15 using your staff email. FAQ: Will old claims still be processed? Yes, if submitted before Aug 15 via the old system. Who do I contact for help? hr@company.com.
Change Announcement and Manager Talking Points
Rolling out an operational change, e.g. a new leave system, to the whole company
no file - description of the change being introduced
- Explain the rationale and the benefit to employees
- Detail exactly what changes
- Prepare candid, no-spin answers to awkward questions
All-staff announcement plus a separate manager briefing sheet
See example AI output
All-staff: we're moving to a new leave system from Oct 1 to fix the approval delays you've flagged. Your leave balance carries over automatically. Manager briefing, Q&A: Why now? System complaints tripled this year, be direct about that. Will my team lose unused leave? No, confirm balances migrate one to one, show the report if asked.
Policy Summary with Compliance Quiz
Turning a long compliance policy into something staff will actually retain
long policy document
- Summarise what employees must do and must not do
- Note who to ask for help
- Write a five-question comprehension quiz with answers
- Stay accurate to the source and keep language plain
One-page summary plus a five-question quiz with answers
See example AI output
Must do: report conflicts of interest within seven days. Must not: accept gifts over RM200 from vendors. Ask: compliance officer, extension 204. Quiz: What's the gift limit? RM200. How many days to report a conflict? Seven days. Who approves exceptions? The compliance officer.
Sensitive Change Messaging
Communicating a difficult change like a restructure or layoffs
no file - description of the difficult change
- Acknowledge the impact honestly and state the facts and timeline
- Explain the support available
- Avoid corporate euphemism and don't overpromise
Written message plus a note on what to say live versus in writing
See example AI output
We're restructuring the Ops team, effective Sept 1. This means four roles will be redundant. We know this is hard news. Affected staff will get one-to-one meetings this week, plus counselling support and two months' severance. Live session: state the number of affected roles verbally first, don't bury it. In writing: keep to facts and next steps, save the empathy for the conversation.
Performance and development
SBI Performance Review Draft
Turning a manager's raw notes into a fair, evidence-based review
manager's raw notes on the employee
- Write two to three strengths using situation-behaviour-impact
- Write two development areas with specific examples
- Set one clear focus for next quarter, no vague adjectives
Written review plus three coaching questions to open the conversation
See example AI output
Strength: in the Q2 client escalation, you stayed on the call past hours to resolve it, which saved the account. Development: in team standups, updates run long, reducing others' airtime, practice a 60-second format. Focus: lead one cross-team project this quarter. Coaching questions: what felt hardest this quarter, where do you want more stretch, what support do you need from me?
Team Rating Calibration Check
Sanity-checking a manager's draft ratings before a calibration meeting
draft performance ratings for the team
- Spot ratings that look inflated or harsh versus the evidence
- Check for uneven standards applied between people
- Flag language showing bias
- Suggest a specific adjustment or question per flag
Calibration review table (Person, Draft rating, Concern, Suggested action)
See example AI output
Person: Ahmad. Draft rating: 5/5. Concern: evidence shows solid, not exceptional, work. Suggested action: ask the manager for one more standout example or adjust to 4. Person: Priya. Draft rating: 3/5. Concern: similar evidence to Ahmad's but a lower score. Suggested action: compare notes side by side, the standard looks uneven.
90-Day SMART Development Plan
Building a concrete growth plan for an employee targeting their next role
employee's strength, growth area, target next role
- Write three SMART goals following a 70-20-10 mix
- Assign on-the-job stretch assignments and recommend learning resources
- Define how progress is measured, with a realistic time commitment
Development plan table plus a manager support note
See example AI output
Goal 1: lead one client presentation solo by day 45, measured by manager observation. 70 percent: shadow then lead two client meetings. 20 percent: weekly feedback from a mentor. 10 percent: complete a three-hour presentation-skills course. Manager note: block two hours a week for coaching check-ins, don't let day-to-day work crowd this out.
Underperformance Conversation Script
Preparing to raise a performance issue directly with an employee
employee's role, the specific performance issue
- Draft an opening that doesn't demotivate, and cite the specific evidence
- Prepare root-cause questions
- Set expectations and close with support and a follow-up date
- Keep the tone firm and kind, using PIP-safe language
Conversation script with branches for a defensive or receptive reaction
See example AI output
Open: I want to talk about the last two project deadlines, I value your work and want to understand what's going on. Evidence: two missed deadlines, dated. Root-cause question: what's been getting in the way? If defensive: I hear you disagree, help me understand your view of what happened. If receptive: thanks for owning that, let's agree two concrete changes and check in in two weeks.
Employee Survey Theme Analysis
Turning open-ended survey comments into a management action plan
open-ended employee-survey responses
- Cluster responses into themes and quantify frequency
- Surface the top three dissatisfaction drivers and top two strengths
- Recommend one action per theme with an owner
- Quote representative comments, don't over-generalise from a few voices
Themed findings table plus a three-action plan
See example AI output
Theme: Workload, 34 percent of mentions, a concern. Quote: constantly firefighting, no time to plan. Theme: Manager support, 12 percent, a strength. Quote: my manager always has my back. Actions: Workload, review headcount against ticket volume, owner Ops Director. Growth, launch internal mobility postings, owner HRBP. Recognition, monthly shoutout ritual, owner team leads.
Employee relations and admin
Grievance Acknowledgement Response
Responding to an employee grievance without admitting liability
the employee's grievance details
- Draft an empathetic, neutral acknowledgement stating the review process and timeline
- Commit to a fair review without pre-judging the outcome
- Keep it factual with no admissions, and flag where to involve legal
Response to the employee plus an internal note on next steps and evidence to gather
See example AI output
Response: thank you for raising this, we take it seriously and will review it fairly. We'll meet with you within five working days and update you on next steps. Internal note: gather the email trail, witness list, and any prior similar complaints. Flag to legal given a potential harassment element before any manager response is given.
Disciplinary Conversation Record
Documenting a disciplinary meeting for the personnel file
raw notes from the disciplinary conversation
- Capture what was discussed and the evidence presented
- Note the employee's response
- List agreed actions and the follow-up date
- Use neutral language and distinguish fact from allegation
Structured record suitable for the personnel file
See example AI output
Date: 12 July 2026. Present: manager, employee, HR. Discussed: three late arrivals in June, dates confirmed per the attendance system. Employee response: cited transport issues, apologised. Agreed action: punctuality to improve immediately, to be reviewed in 30 days. Follow-up date: 12 August 2026.
Engagement Pulse Survey
Launching a quick, recurring pulse check on employee sentiment
no file - generic for the organisation
- Write two questions each for workload, manager support, growth, recognition and belonging
- Apply a consistent five-point scale plus one open comment question
- Keep wording unbiased and completion time under three minutes
Survey with its scale plus a note on how to analyse the results
See example AI output
My workload is manageable, one to five. My manager supports my development, one to five. I feel I belong here, one to five. Open comment: what's one thing we could do better? Analysis note: segment by department and tenure, flag any category averaging below 3.5 for follow-up.
Complete Offboarding Pack
Preparing everything needed when an employee is leaving
the departing employee's role
- Write a farewell email
- Draft eight exit-interview questions for honest feedback
- Build a knowledge-handover checklist
- Keep the tone dignified regardless of the reason for leaving
Farewell email, exit-interview questions, and handover checklist
See example AI output
Farewell email: thank you for three great years, you'll be missed. Exit questions: what made you decide to leave, what would have kept you, how would you describe your manager relationship? Handover checklist: transfer client files, document open tickets, revoke system access on the day of exit, brief the successor by the agreed date.
Monthly People Dashboard Summary
Reporting headcount, turnover and leave trends to management each month
headcount, turnover and leave data
- Summarise headcount movement
- Break down voluntary versus involuntary turnover and the trend
- Flag two risks worth management attention with a suggested action
- Use only the attached data and state any data gap
One-page summary highlighting the three numbers that matter most
See example AI output
Headcount: 142, up 3 this month. Voluntary turnover: 2.8 percent, up from 1.9 percent. Involuntary: 0.7 percent. Top three numbers: 142 headcount, 2.8 percent voluntary turnover, 68 percent average leave utilisation. Risk 1: voluntary turnover rising in Ops, recommend stay interviews. Risk 2: low leave utilisation may signal burnout. Gap: no exit-interview data attached this month.
More prompt packs by function
View the full Prompt LibraryFrequently Asked Questions
Because each one gives the AI a role, tells it what you will attach (the CV, the policy, the review notes), specifies the framework to follow, and defines the output. That produces fair, structured, defensible HR work you can use, rather than a generic paragraph you have to rewrite.
Only to your organisation's approved, enterprise AI where data stays in your tenant. Never paste names, salaries, IC numbers or grievance details into public consumer tools. HR data is sensitive under Malaysia's PDPA — anonymise where you can and follow your data policy.
It can speed up first-pass screening against your criteria, but you must guard against bias and keep a human in the loop for decisions. These prompts deliberately instruct the AI to judge only on the stated criteria, flag bias-prone inferences and never auto-reject.
Copilot is convenient if HR runs on Microsoft 365; Claude and ChatGPT are strong for drafting and structured evaluation over uploaded files. These prompts work across all of them.
Yes. Our AI courses are HRD Corp SBL-KHAS claimable for eligible Malaysian employers, so HR teams can upskill at near-zero net cost.
Yes. Attach performance ratings and potential assessments, anonymised where possible, and ask AI to plot employees onto a 9-box grid of performance versus potential, then suggest development actions for each box, for example stretch assignments for high-potential, high-performance staff, or a structured improvement plan for low performers. AI is strong at organising the data and drafting the narrative, but the actual placement decisions should be calibrated by a panel of managers, not the AI alone, since succession calls carry real career consequences and benefit from multiple human perspectives.
AI is useful for drafting plain-language explainers of EPF, SOCSO and EIS contribution rates, eligibility and claim processes that employees can self-serve, cutting repetitive HR queries. Ask it to turn your current statutory rate table into a simple FAQ or chatbot script. Always verify the rates against the latest KWSP, PERKESO and LHDN publications before publishing, since statutory rates and thresholds change periodically and AI training data can lag official updates. Treat the AI output as a first draft for your team to check, not the authoritative source of truth.
AI can analyse patterns you already track, such as engagement survey scores, tenure, overtime and manager change frequency, to flag employees showing common early warning signs of resignation, and suggest a tailored retention conversation for each. It works best as a triage tool that prompts where a manager should have a genuine check-in, not as a predictive verdict on any individual. Because flight-risk scoring touches sensitive personal inferences, keep the output internal, use it only to prompt supportive conversations, and never let a score alone drive a decision about an employee.
Give the AI the employee's role, the specific ground such as redundancy, poor performance or contract non-renewal, the notice period per the contract or the Employment Act 1955, and any prior documentation such as performance improvement plan records. Ask it to draft a factual, respectful letter stating the reason, effective date, final entitlements and next steps, while flagging any clause that needs legal review. Termination carries high legal risk in Malaysia, so always have the final letter reviewed by an employment lawyer before it is sent, since AI drafts the structure but does not replace legal sign-off.
Yes, attach your anonymised headcount, hiring and promotion data broken down by gender, age band or other tracked categories, and ask AI to calculate representation ratios, flag gaps against your targets, and draft the narrative section of a DEI report. This saves hours of manual data wrangling. Keep the underlying data anonymised or aggregated before it reaches any AI tool, since individual-level demographic data is especially sensitive under PDPA, and have a human review the framing carefully before the report goes to management or a regulator.
Go beyond prompts — train your team
Prompts are the start. AITraining2U runs hands-on, HRD Corp SBL-KHAS claimable AI training for Malaysian teams — from everyday AI productivity to building agents that run human resources workflows end-to-end.