Before you begin
Start with one workflow, one source of information, and a named reviewer.
Use sanitized sample data while you are testing the structure and review loop.
Build the workflow
Matching is a strong first AI workflow because the result is easy to inspect.
The app receives
A person, company, request, or project
A set of candidates
Required criteria
Preferred criteria
Disqualifying conditions
It returns
Ranked matches
Score by criterion
Reasons for the ranking
Missing information
A reviewer correction
Starter prompt
“Build a matching app for [user] to match [request] with [candidate type]. Apply [required criteria] first, then rank by [preferred criteria]. Explain every score. Do not recommend a candidate that fails [disqualifying rule]. Let the user correct the match.”
Start with a small candidate set and a visible scoring table. Do not hide the reasoning inside a single overall score.
Common mistakes
Hiding the reasoning inside one overall score.
Allowing preferred criteria to override a disqualifying rule.
Failing to capture missing information and reviewer corrections.
Frequently asked questions
Can I use this template for a different industry?
Yes. Replace the role, input, rule set, output, reviewer, and prohibited actions. Keep the first workflow bounded and inspectable.
When should I use production data?
After the structure works with safe examples and the required permissions, security, retention, and review controls are in place.