Why I Built My Own AI Consultant (And Why You Should Too)
I'm Matthew Stublefield, and I help product professionals develop the leadership skills they need to advance their careers. Recently, I spent $2,000 on a course that includes a custom AI tool designed to help me build better courses.
The irony? Within 20 minutes of using it, I realized I could build something better myself in about 5 minutes.
Let me explain.
The Premium Course With a Limited AI Tool
I'm taking Amy Porterfield's Digital Course Academy (fantastic program, by the way—I'm genuinely excited about what I'm learning). One of the new features this year is "Porter," an AI assistant loaded with all of Amy's training materials to help you plan and build your digital course.
Here's what I discovered: Porter is locked to specific modules, doesn't save your progress, can't access any information about YOUR business, and you can't customize it. It's helpful in a generic way, but it doesn't know anything about my positioning, my customers, or what makes Fieldway different.
The Problem With Pre-Built AI Tools
This is the problem with most pre-built AI tools.
I've bought GPTs for $7, $17, even $27. At first, they seem great. But you quickly hit the walls: you can't modify them, can't give them your business context, can't adapt them to your actual workflow. You're stuck with someone else's 60% solution while you manually fill the gaps.
So I rebuilt my instructional designer AI using Claude.
Took me maybe 10 minutes total.
And it's dramatically better because it knows:
My positioning documents
My ideal customer challenges
My teaching frameworks
My business context
The Simple 4-Step Process
Here's the process I used (and teach in my new course):
Start with a basic prompt describing what you need
Ask the AI to ask YOU questions to understand your context better
Have it generate research prompts for Perplexity
Review the research to make sure its legit (this is still necessary and will be for a while)
Bring the research back and build your custom instruction set
(Want this process with more detail in a snazzy infographic? I got you.)
That's it. No coding. No programming knowledge. Just a systematic approach that takes a few minutes once you know the process.
The result? AI consultants that give you specific, contextualized advice instead of generic recommendations anyone could get.
Try This Technique Next Week
Open Claude or ChatGPT and describe something you need help with. Then ask it: "What questions do you need to ask me to provide better, more specific guidance?"
See what happens. This simple meta-prompting technique is the foundation of building AI consultants that actually understand your business.
Who Should Build Their Own AI Consultants?
This approach is perfect for:
Product professionals who need expert guidance across multiple domains
Entrepreneurs tired of paying for tools that almost work
Business owners who want AI that understands their specific context
Anyone frustrated with generic AI responses that could apply to any business
Stop Adapting to Generic Tools
The bottom line: Stop adapting your business to someone else's AI tool. Learn to build AI consultants that adapt to YOUR business.
If you want to learn the exact process I use, I created a 2-hour course called Build Your AI Consultant. It's $47 (or 3 payments of $19), and includes the worksheets, prompts, and frameworks I actually use.