Use case B7 · Business
Hire with AI's help, from the job post to the shortlist to grading interviews: a software founder and an agency owner
- Hiring
- Beginner
- ChatGPT, HubSpot, Claude
- Source checked · LinkedIn post
The short version
What they did
Two owners used AI to run most of a hiring process: writing the job post, sorting a pile of applicants down to a shortlist, and grading interviews and test tasks. In both, the owner made the final call.
A software founder’s first hire
Nicole Pereira had ChatGPT write the job description, compare hiring regions, write recorded-interview questions with a grading matrix, and design a paid test. HubSpot AI summarized the top 20 applicants. Her LinkedIn post, 2025-06-27.
An agency owner’s shortlist
Andrius Avinas of Adguns refined the job post with AI, then fed it CVs, motivation letters and personality-test results with the role context. He reviewed recorded intro-call transcripts with it, and built and graded a test task. His LinkedIn post, 2026-06-16.
Result (as claimed)
The software founder: “I posted a job that had almost 400 applicants in less than 3 days. And Ai did most of the work.” “I now have a responsive and qualified applicant taking my paid test and we are having our second interview on Tuesday!”
The agency owner: “Out of 20 candidates, AI helped highlight the 5 strongest fits with solid insights layered on top.” “AI kept surfacing the same strong candidates that my team also flagged.” “The result was really strong. We’ve got a new hire.”
Prompt
The job post and interview kit:
I’m hiring a [role] for my [business or product]. The work, hours, location, pay range and required skills are [paste]. What good performance looks like is [describe]. Our hiring process and tone are [describe].
1. Draft a job post under 400 words using only the requirements I provide.
2. Create 10 recorded-interview questions and a short invitation email. Give a scoring sheet tied to four or five job skills, each scored from 1 to 5 with clear meanings.
3. Design a short paid test with instructions, a suggested time limit and a matching scoring guide.Use plain, welcoming language. Ask about missing pay or duties rather than inventing them. Keep questions and scores tied to the job, without using protected personal traits. I will review the materials and make the hiring decision.
The shortlist:
I’m reviewing applicants for [role] at my [business]. The duties, required skills and scoring rules are [paste]. The candidate materials are [attach CVs, letters and relevant test results], each labeled with a candidate ID. Help me prepare a shortlist for human review.
1. Score each candidate from 1 to 5 against the same job requirements. Quote evidence from their materials and identify its source.
2. List unanswered questions to probe in an intro call. Treat missing information as unknown, not proof of poor ability.
3. Suggest up to five candidates to review first, explaining the order and any ties.Return a comparison table and a brief note for each candidate. Don’t invent experience or infer protected traits, personality or ability from names or background. Flag unsupported test interpretations. I make the final call.
Do it yourself
These steps are my suggested way in, not any one person’s setup.
What you need
- A chat assistant such as ChatGPT or Claude.
- A clear picture of the role: the tasks, the hours, the pay and what good looks like.
- Applicant materials you are allowed to share, such as CVs and cover letters.
Start simple
- Write five lines about the role in your own words. Paste them with the first prompt above.
- Edit the job post it drafts. Cut anything you would not say out loud to a candidate.
- Ask for a short scoring sheet tied to the role: four or five skills, each scored 1 to 5.
- When applications arrive, paste a few at a time with the second prompt. Ask for quoted evidence for each score.
- Read every application on the shortlist yourself before you contact anyone.
Take it further
- Add a short paid test task with a rubric, and have the AI grade it against that rubric only.
- Use transcripts of recorded intro calls and ask what to probe in the next interview.
- Keep one saved project with the role context so every candidate gets judged the same way.
Check your work
- Spot-check a few candidates the AI ranked low. Make sure it did not skip someone good.
- Check every quote it uses against the actual CV. It can invent or blend details.
- Hiring is regulated. Tell it to ignore age, gender, race, disability and other protected traits, never let it make the final call, and check your local rules on using AI in hiring.
Source checked. The link, author, date and quote were confirmed against the source. The result itself has not been reproduced. Prompt adapted for this library, inspired by the case. Business size: Solo founder (first hire); agency with production managers. Results are quoted as the source claimed them and never improved. Spot an error or a broken link? Email charlie@cbarmorecpa.com.