Steal My AI Prompting Process

For 12 years, I’ve struggled to answer one simple question.

What do you do for work?

I work in a niche that’s hard to explain quickly.
Staffing and recruiting is already not known by most people.
Automation and AI inside staffing and recruiting even more so.

I never had a good answer.
Even family and friends only have vague ideas of what I do.

I work with technology.
I travel sometimes.
I speak at events.
I help people.
I work remotely.

All true.
Still confusing.

Yesterday I finally solved it and AI helped me do it.

Not because AI magically wrote the perfect sentence on the first try.
Because I used a framework that makes AI the ultimate assistant.

Most people use AI like this:

Write me a better intro.
Then they accept whatever comes back, even if it doesn’t match their voice, their audience, or the vibe they want.

The solution is simple.

Do not ask AI for the final answer.

Ask AI for options, evaluation of those options, and improvement.

Here’s the full step by step process of what I did.

This is the exact loop I used.

1. Produce options

2. Evaluate one option against clear criteria

3. Generate better alternatives

4. Combine what I like

5. Repeat until I’m happy with it

Below is how to do each step.

Your goal needs to be specific, and your constraints need to be explicit.

I told AI what I wanted and what I didn’t want.

First prompt I used.

Prompt 1
I’m an industry expert in Automation and AI for Staffing and Recruiting. How do I quickly explain to someone I just met what I do if they don’t know anything about Automation and AI or Staffing and Recruiting? I want something clear, non technical, explains my expertise, and is easily expandable in further conversation if they are curious. Give me 5 options.

Key principles:
- Ask for multiple options.
- Define what you want
- Define what you don’t want

Out of the five, the one I liked most was:

Staffing companies help people find jobs and companies find people. I’m the best in the world at helping staffing companies with Automation and AI improve their outcomes.

Good start.
But it had issues.

Best in the world is INSANELY braggy.
It didn’t clearly communicate impact.
It wasn’t phrased the way I talk.

So instead of trying to fix it myself, I had AI evaluate it.

You give AI an example, then you give it a scoring rubric.

Prompt 2
Staffing companies help people find jobs and companies find people. I’m the best in the world at helping staffing companies with Automation and AI improve their outcomes.

Evaluate the above as an introduction to someone who has no clue about any of it.

Evaluation goals:
Clear
Non technical
Non braggy
Memorable
Expandable if they’re curious

Give me 3 more optimal options.

This works unbelievably well because now AI isn’t guessing what you want.
It’s grading against your rules.

The next three options it gave me included phrases I loved:

Fill jobs faster
Make more placements
Remove bottlenecks
Increase revenue without adding headcount
Automate the repetitive parts

I didn’t pick one.

I combined the best parts into a new version.

Now I had a new draft, so I ran the same evaluation again.

Prompt 3
Evaluate this one in the same way with the same output:

Staffing companies help businesses hire people and people find jobs. I help those staffing companies use automation and AI to increase revenue, provide a better experience, without increasing their headcount.

AI responded with better versions.
I selected what I liked and repeated the loop.

We were heading in the right direction but it felt like I lost the signal that I’m good at what I do.

So I asked for that specifically.

Prompt 4
Give me a version that subtly signals I’m good at what I do without saying it.

That produced versions like:

I work with staffing leadership teams
I help executive teams identify operational bottlenecks
I design systems that create measurable leverage

Those phrases imply seniority and expertise without ever saying best or leading.

After a few more combinations and further evaluations, I ended here:

Staffing companies help businesses hire top talent. I work with staffing executive teams to design and implement automation and AI systems that increase revenue and improve candidate and client experience without adding headcount.

Then I stopped.

That is another key principle: stop when it sounds like you and you like it.

My wife tested it in a conversation and said, for the first time, she actually knew how to explain what I do.

That’s when I knew I was successful.

The difference between average output and optimal output isn’t the tool.

It’s the process:

Produce.
Evaluate.
Combine.
Refine.
Repeat.

That process works for me.
I’m willing to wager it’ll work for you too.