From mind map to app: How anyone can build with vibe coding
Hannah
Vibe coding—building apps by describing what you want in plain language and letting AI generate the code—was named Collins Dictionary's Word of the Year for 2025. And according to recent industry data, 63% of people doing it today aren't developers at all. They're teachers, retailers, small business owners, and anyone with a problem worth solving.
"I don't know how to code. I never have," Suleiman Shaibu told the small group gathered on Xmind's first webinar of 2026. "But I've built over 10 apps. And today I'm going to show you how."
Suleiman is a management consultant and business coach based in Nigeria, with over 30 years of experience working with companies like Microsoft and HP. He discovered mind mapping almost a decade ago—and when AI came along, he realized the two belonged together.
This session wasn't a technical tutorial. It was a demonstration: that vibe coding isn't reserved for developers. It's for bakers, teachers, retailers, and anyone who has ever had a problem worth solving.
The architect analogy: Why most people get vibe coding wrong
Before touching any AI tool, Suleiman made one thing clear: the way most people approach vibe coding is exactly the wrong way.
"You just don't go to a construction site and say 'build me a house,'" he said. "Even the smartest architect will ask you questions first. How many rooms? How many floors? Where do you want the kitchen?"
AI is your builder. But without a blueprint, it's just guessing. That blueprint, Suleiman argued, is your mind map—and this is where Xmind fits into the workflow. Not as a finishing touch, but as the starting point.
Before writing a single prompt, he maps out:
The problem — What exactly are you trying to solve?
The users — Who will actually use this, and how?
The features — What does it need to do?
The flow — What happens when someone opens the app?
The map becomes your product requirements document. The clearer it is, the sharper the prompt. The sharper the prompt, the closer the first version gets to what you actually had in mind.
Building live: The energy recharge tracker
To make the idea tangible, Suleiman built a real app during the session—starting from scratch, in real time.
Back home in Nigeria, his family had been manually tracking their prepaid electricity top-ups: the date, the amount spent, the meter reading at the time. Anyone in the house could do a recharge, but keeping a shared record was messy. So he opened Xmind and started thinking it through out loud—mapping out users, login options, and core features before touching any AI tool.
Category | Details |
|---|---|
Users | Father, mother, son, daughter, occasional guest |
Login | Google login + username/password (for guests without Google accounts) |
Core features | Log recharge date, amount, meter reading; view history |
Analytics | Weekly/monthly usage reports, recharge history |
As the map took shape, attendees jumped in with suggestions—multi-currency support, two-week recharge history, longer reporting periods. Suleiman added each one directly to the map before writing a single prompt.
"The good thing about mind mapping is that once you start, your creative juice starts to flow," he said. "Things you never considered start to show up."
With the map complete, he fed it into 3 AI tools simultaneously—Lovable, Google AI Studio, and Claude—to compare outputs. All 3 returned working mockups within minutes. Google AI Studio was first, and had already picked up on the Nigerian context: it defaulted to naira as the currency without being asked.
The first versions weren't perfect. Buttons didn't respond. Currency options didn't update. But that's exactly the point.
"AI, like human beings, makes mistakes," Suleiman said. "That's why we have iteration. You keep refining the instruction until you get what you want."
The 7-step vibe coding framework
Over the course of building more than 10 apps, Suleiman distilled the process into a repeatable framework:
Define the problem — What are you solving, and for whom?
Identify your users — Who will actually use this?
Brain dump features — Map everything in Xmind, then use Xmind AI to surface what you might have missed
Prioritize core features — Build a minimum viable product first; layers come later
Generate your first version — Feed the map to your AI tool of choice
Iterate — Fix, refine, and expand based on what comes out
Deploy — Host on platforms like Netlify or Vercel
One tip surfaced from an attendee mid-session: instead of screenshotting your Xmind map and attaching it as an image, export it as Markdown or Word and paste the text directly into your prompt. The AI reads structure better than it interprets visuals, and the output tends to be sharper.
"Well done. Very good suggestion," Suleiman responded. It was one of those moments that made the small group format feel like an advantage.
Same process, different problems
To show the framework wasn't industry-specific, Suleiman walked through 3 more use cases in quick succession.
Bakery inventory app
A baker needs to track ingredients, manage stock, and know when to reorder. Suleiman mapped the features in Xmind and handed them to Google AI Studio. The result was a functional inventory app—and it included something he hadn't thought to ask for: a built-in AI assistant that lets you query stock in plain language instead of navigating tables.
He then showed how a few more prompts could extend it further:
Bulk CSV upload for new inventory
Automatic cost price updates from supplier sources
Margin tracking across products
Each new feature was one prompt away.
Clothing boutique sales app
A retailer struggles to remember returning customers' preferences—size, color, fabric. Suleiman mapped the problem in Xmind and sent it to ChatGPT, which came back with something unexpected: full UI mockup suggestions that raised design questions his original map hadn't considered.
"This is why we need to brainstorm," he said. "There's no way your AI would know what's in your mind. So before you jump in, map it out as completely as you can."
The resulting app stored customer profiles, logged purchase history by preference, tracked inventory, and could send promotional emails to returning customers.
Networking icebreaker app
Perhaps the most personal example. Suleiman built this one for his own facilitation work—an app that generates icebreaker question sets for corporate retreats, runs them as a timed game via QR code, and tracks responses. No template. No tutorial. Just a problem, a mind map, and a prompt.
Conclusion
Toward the end of the session, one attendee mentioned something Elon Musk had said recently—that the future of phones might be a single device with no fixed apps, where everything is generated on demand. After watching Suleiman build 4 working apps in under an hour, that idea felt a lot less abstract.
His closing message was straightforward.
"You cannot afford to be helpless in this new age. Your future is in your hands. You decide what's important to you."
The tools are there. The barrier isn't technical anymore—it's clarity. Vibe coding gives you the builder. Xmind gives you the blueprint. What you build is entirely up to you.
FAQ
1. What is vibe coding?
Vibe coding is a way of building apps by describing what you want in plain language and letting AI generate the code. The term was coined by AI researcher Andrej Karpathy in early 2025 and named Collins Dictionary's Word of the Year for 2025.
2. Do I need to know how to code to start vibe coding?
No. Research shows that 63% of people currently vibe coding are non-developers. The skill you need isn't syntax—it's clarity. The better you can describe what you want, the better the AI performs.
3. What is the best way to prepare before using a vibe coding tool?
Map out your idea first. Define the problem, identify your users, and list the features you need. Tools like Xmind help you organize your thinking visually before you write a single prompt—which leads to significantly better results.
4. Can I really build a working app without any technical background?
Yes—with the right approach. As Suleman demonstrated live in this webinar, a clear mind map translates directly into a working app prompt. The AI handles the code; you handle the thinking.
5. Which AI tools work best for vibe coding?
Popular options include Lovable, Google AI Studio, Claude, and ChatGPT. Each has different strengths—it's worth testing a few with the same prompt to compare outputs and pick the best starting point.
