Workforce Wonkery bee logo

Workforce Wonkery

Policy. Data. Practice. Decoded.

AI Readiness Lab · Job Search Practice

Build a better job-search prompt

Use a fictional candidate and a synthetic job posting to learn how AI can strengthen an application without inventing the person.
About 3 minutes · No participant information required
Meet Alex

Start with evidence the candidate actually has

Alex is fictional. The job posting is synthetic. Everything you need is already on the page so the learning stays focused on prompt design.

Alex’s experience

Alex has six years of retail-supervision experience. They schedule a 12-person team, track inventory, coordinate incoming deliveries, train new staff, use spreadsheets for weekly reports, and handle escalated customer issues. Alex wants a weekday role with less constant customer contact. They have not managed a budget, negotiated contracts, or supervised 25 people.

Real evidence
Scheduling, inventory, deliveries, training, spreadsheets
Goal
Move into operations coordination
Preference
Weekday schedule, less customer-facing work
Important boundary
No budget, contract-negotiation, or 25-person supervision experience
Synthetic posting · Operations Coordinator
  • Coordinate schedules and daily workflow.
  • Track inventory and incoming materials.
  • Communicate with vendors and internal teams.
  • Maintain spreadsheets, reports, and operating records.
  • Support onboarding and staff training.
A weak prompt

“Rewrite Alex’s resume for this Operations Coordinator job.”

Build the prompt

Add four things that protect quality

Click each ingredient and watch the prompt improve. The structure can be reused later with a real candidate using your organization’s approved AI environment and privacy rules.

Your prompt0 of 4 ingredients added
Copies the prompt to your clipboard. Nothing is sent from Workforce Wonkery.
The reusable pattern

For job-search work, a strong prompt should identify the target, provide real evidence, set truthfulness boundaries, and define the output you actually need.

Judge the answer

Polished language can still be false

Imagine AI returns this resume bullet. What would you do?

Sample AI output · intentionally imperfect

“Managed a 25-person operations team, controlled a $500,000 annual budget, and negotiated vendor contracts to improve inventory performance.”

What you learned

AI should strengthen evidence, not manufacture it

TargetWhat job and requirements are we responding to?
EvidenceWhat has the candidate actually done?
BoundariesWhat must AI never invent or exaggerate?
OutputWhat useful work should AI produce?
Then use human judgment

Check that the wording is truthful, recognizable to the candidate, relevant to the posting, and something the person could comfortably explain in an interview.

Next in the path

You have practiced protecting the truth. Next, use AI for decision support without handing a consequential decision to the model.

Practice with fictional or appropriately de-identified information. Public job postings can be useful context; participant information requires your organization’s approved AI, privacy, security, and records practices. The candidate remains responsible for the truth of the application, and the counselor remains responsible for professional judgment.