SHRM developed this guide using the same general-purpose AI tools it describes. The prompts are intended as starting points and may need to be adapted, reviewed, and validated for your specific use case.
AI can help you get more done with the HR work you already do, from drafting and summarizing to organizing information and generating ideas. But it does not know your organization, your people, your policies, or your obligations unless you provide that context. This prompt gives you a starting point, with space to add the unique information AI needs to support your work.
Prompt 1 — Audit Your Candidate Records
Run this before any cleanup. It tells you what kind of duplication you have, which determines whether this is a two-hour job or a two-week one.
I have an export of candidate records from our applicant tracking system. Before we clean anything, help me understand what is wrong with it.
[PASTE A SAMPLE OF 100-200 RECORDS — REMOVE EMAIL ADDRESSES, PHONE NUMBERS, AND ADDRESSES; KEEP IDS, NAMES, DATES, AND APPLICATION HISTORY]
Tell me:
- What kinds of duplication you can see — name variants, re-applications, records created by different sources, test or junk records.
- Which fields are inconsistently formatted and how.
- Roughly what proportion of these records look like they have at least one duplicate.
- Any field that is empty often enough to be unreliable for matching.
- What additional field, if we had it, would make matching much more accurate.
Do not merge or recommend deleting anything yet. I want a diagnosis.
Prompt 2 — Flag Potential Duplicate Candidate Records
Replace everything in brackets before you run the prompt. Optional, if available: give the AI your organization's style guide, brand voice guide, or writing standards document alongside this prompt so the output matches how you write.
Identify likely duplicate candidate records in this export.
[PASTE RECORDS — IDS, NAMES, APPLICATION DATES, REQUISITIONS. NO EMAIL ADDRESSES, PHONE NUMBERS, OR ADDRESSES.]
For each group of records you believe belong to one person:
- List the record IDs.
- Give a confidence level: high, medium, or low.
- Explain exactly what made you group them — name variant, shared history, timing pattern.
- Note which record appears to be the most complete and why.
Rules:
- Do not merge anything. Produce a review queue only.
- Treat a re-application to a different role at a different time as one person with a history, not as an error to remove.
- Never group records on the basis of name origin, ethnicity, or anything you infer about the person.
- List separately any record that looks like test data or junk.
- Flag any grouping where you would want a human to look closely before acting.
Write in plain, everyday language. Do not use jargon or invented terms; if a simpler word exists, use it.
Check Before You Publish
Before you publish, review the AI-generated draft carefully. You are responsible for the final version.
- Check facts, figures, dates, requirements, and other details against the information you supplied or the original source.
- Paste in any legally required or compliance-sensitive language yourself. Do not let AI generate it.
- Confirm nothing was invented, assumed, or taken out of context.
- Verify that the final version reflects your organization, your audience, and the purpose of the work.
Use your judgement as an HR professional. AI can assist with the work, but you make the final call. Looking for more prompts? Explore the AI Prompts Guide for HR for more prompts and templates to support your HR work.
Combine AI with Human Intelligence
Explore SHRM’s AI+HI Specialty Credential to learn how to use AI strategically while applying the human judgment HR work requires.
Was this resource helpful?
Validate your HR expertise
Earning your SHRM-CP credential makes you a recognized expert and leader in the HR field.
Related Content
Learn how Marsh McLennan successfully boosts staff well-being with digital tools, improving productivity and work satisfaction for more than 20,000 employees.
The proliferation of artificial intelligence in the workplace, and the ensuing expected increase in productivity and efficiency, could help usher in the four-day workweek, some experts predict.
As artificial intelligence technology continues to develop, the demand for workers with the ability to work alongside and manage AI systems will increase. This means that workers who are not able to adapt and learn these new skills will be left behind in the job market.