From Zero to App Builder: Dimitri Built the Tool
Jan 15, 2026Dimitri had a familiar kind of problem: a good idea with nowhere to go.
He's been in real estate for more than a decade. Long enough to know the difference between "interesting data" and "actionable signal." Long enough to watch people make expensive decisions off vibes, gut, and spreadsheets that slowly evolve into abstract art.
He wanted a market analysis tool that could do two things cleanly: First, compare a property against government assessment data and output a sensible value score. Second, surface market pressure indicators without making you read a 40-tab dashboard like it's sacred scripture.
The idea wasn't hard. The build was the wall.
He couldn't code. Not in the "I watched a tutorial once" way. In the "I have a business and a life and I'm not doing a second unpaid degree" way.
The Options That Weren't Options
Hiring devs is a tax on clarity. You pay money, then pay again in translation errors. Features get lost. Timelines expand. Ownership gets weird.
Learning to code is great if you want a new career. Dimitri wanted a tool.
No-code is fine until you need a custom thing... and then you're back to duct tape and prayer.
So the concept sat on the shelf. Not because he wasn't serious. Because the system was built to keep builders and domain experts in different rooms.
In November 2025, Dimitri joined Cohort 7 of AI Upgrade for Creative Professionals. That's when the rooms started connecting.
The Real Shift: Software Stopped Caring Who Writes the Syntax
Here's the thing that's quietly wrecking old assumptions:
The bottleneck isn't "can you write code" like it used to be.
The bottleneck is "can you describe the intent with enough precision that a machine can help you execute."
Syntax mattered because it was the gate. Now semantics is the gate.
Tools like Lovable, Bolt, and v0.dev didn't eliminate engineering. They changed where the effort goes. Instead of spending hours proving you can type curly braces, you spend hours making decisions: what the product is, what it does first, what users need to see, what "good" looks like.
That's not less work. It's better work.
And it favors people like Dimitri. People with actual domain muscle.
Dimitri had that. Real estate teaches you fast.
The Unsexy Step That Made Everything Possible
Before he "built an app," Dimitri did the thing most people skip: he organized his thinking.
Notion became the staging area. Not as a productivity cosplay. As a way to take years of mental notes and turn them into something buildable: feature list, data requirements, scoring logic, UI priorities, edge cases.
Then Peter taught the POP framework: Persona, Objective, Parameters. It's basically how you avoid prompting like you're summoning a genie.
Persona: who the AI should be.
Objective: what outcome you want.
Parameters: constraints, style, what to avoid, what matters.
Dimitri started prompting like he was briefing a competent teammate instead of fishing for miracles. The quality jumped. Not because the model got smarter—because the instructions did.
Another important thing: don't overcomplicate it with "agents"
We also hit Peter's Agentic Spectrum: assistants, automations, agents.
Dimitri didn't need an autonomous agent doing weird stuff unattended. He needed an assistant workflow: he drives, the tool drafts, he approves, repeat.
That decision matters. A lot of people think the future is "let the agent run everything." Usually it's "let the assistant do the boring parts while you stay accountable." That's how you build something you'd trust with real decisions.
The Moment It Turned Into a Thing
During the Vibe Coding workshop we used Lovable.
Dimitri described the product in plain language: input property data, compare to government assessments, produce a score, show market pressure indicators, keep it clean and usable.
Lovable generated a working prototype.
Not a slide deck. Not a clickable Figma illusion. A deployed app-shaped object you can interact with.
What He Built, in Real Terms
The first version was simple and honest: basic interface, scoring, government data coming in via upload, indicators that communicated something without being precious.
Then he iterated.
He tightened the UI based on feedback. He learned to prompt in batches so he wasn't burning credits on repetitive micro-changes. He made it mobile-responsive because if it breaks on a phone, it's not a tool—it's a hobby.
Then he added market pressure indicators and location-aware insights. He started planning API connections to live listing data.
He didn't spend time "coding." He spent time describing, testing, refining. Roughly 10–15 hours.
That's the trade: less time wrestling syntax, more time wrestling truth.
Demo Day: The Only Kind of Proof That Counts
On capstone day, Dimitri showed the cohort the tool. It did what it said it would do.
The best reaction wasn't applause. It was people asking if they could use it.
The Actual Lesson (Not the Inspirational Poster Version)
Dimitri didn't become a programmer.
He became an app builder because he already had what matters: a clear problem, lived expertise, and the willingness to iterate in public.
The AI didn't hand him a business. It handed him a faster feedback loop.
And that loop is the power:
Build the smallest real version.
Show a human.
Change it immediately.
Repeat.
Shipping isn't the end. Shipping is when reality finally starts contributing.
Where He Goes Next
Now he's thinking about the next layer: live listing data, user accounts, saved searches, advanced filters, alerts, maybe a mobile-first version.
The interesting part isn't the feature list. The interesting part is that he's not waiting for permission anymore.
He's not stuck in "someday." He's in "next."
The Bigger Picture
For a long time, software creation was a closed room. If you didn't speak code, you stood outside and hoped someone translated your idea faithfully.
That room is opening.
Not because machines are magic. Because the interface for creation is shifting toward language, intent, and iteration.
So yeah, the question isn't "can you code."
The question is: do you know what you're trying to build, and can you explain it clearly enough to make it real?
Dimitri did.
Now he has a tool. And a new baseline for what he's capable of.
Ready to Build the Tool You've Been Thinking About?
Join the next cohort of AI Upgrade for Creative Professionals. Six weeks. Live sessions. Real coaching. Actual results.
The future's not coming. It's already here. We're just building it one capstone at a time.
Kris Krüg is co-founder of The Upgrade Academy, recovering photographer, and professional instigator of creative chaos. He's spent 20+ years watching technology either amplify or stifle human potential. He's chosen amplification.
Peter Bittner entrepreneur, AI product thinker, new media journalist, and UC Berkeley lecturer, is the methodical yin to Kris's creative yang, turning chaos into deployable systems. Together, they've empowered hundreds of professionals.
