I Built An App with AI in 60 Days. How Much Money Did it Make?

Product Reviewed: Flow Stack

Flow Stack is a “link-in-bio” one-page website builder for creatives (musicians, DJs, photographers, videographers, content creators, etc.). Users can create a page that includes:

• Links
• Photos/videos
• Media galleries
• Upcoming events
• A live preview shown in a dashboard

Key Features Mentioned

• One-page website builder (“link in bio” style)
• Dashboard to upload and manage content (links, photos, videos, media)
• Live phone preview updating in real time as the site is built
• Authentication system: sign up, log in, log out
• Subscriptions / payments, with multiple billing flow options considered:
  • $3/month billed yearly (stated as “yearly subscription, $3 per month”)
  • $4/month billed monthly
  • Consideration of variants like free trial and credit card timing options
• Security & data handling
  • Stores user data (email/password + uploaded media/files) in Supabase
  • Includes goals around stability and security error handling

Build Experience / User Experience (Reviewer POV)

• The reviewer is not a developer and began building with about ~2 months of work initially (later extended).
• They used AI “vibe coding” via Lovable (their best tool choice based on research) and ChatGPT prompts.
• Back-end generation worked well, including auth and database/data flow.
• The front end initially looked “terrible/basic”, including:
  • A basic logo and dark mode
  • Missing or incorrect core UI flows (e.g., no pages and no working login)

Manual front-end work

The reviewer had to:

• Manually design most UI in Figma/Canva
• Then guide the AI using screenshots/reference images so the AI could replicate the intended designs

Main challenge: dashboard + live preview reliability

The hardest part was making dashboard + live phone preview work reliably:

• New AI-generated features sometimes broke the app or broke the preview
• Required multiple prompt/fix rounds over about two weeks
• Reported login bugs (sign-in worked briefly, then immediately logged out), requiring heavy debugging

They eventually reached a stable, end-to-end working app, then spent additional time addressing security issues.

Pros

• Fast initial app creation with AI for non-developers (especially back end)
• Useful functionality delivered, including sign-up/login, content management, logout, and payments
• Low operating cost
  • $20/month for Lovable
  • $25/month for Framer (landing page)
  • Total: $45/month
• Break-even math claimed
  • At $4/month pricing, they estimate ~11 users/month to break even

Cons / Limitations

• Front-end still needs human oversight
  • AI struggles with UI detail and consistency
  • Even small UI constraints (e.g., “don’t change this menu background”) repeatedly failed
• Unreliable integration while iterating
  • Adding features often broke existing components, especially dashboard/live preview
• Security/data reliability is a major risk
  • Since the app stores sensitive user data, Supabase issues could create stability and security risks at scale
• Time/effort still significant
  • Reported ~380 hours and $326 invested to reach a working, secure state

Comparisons Made

• Compared the target to similar tools Lingree and Comey, stating they lacked features the reviewer specifically wanted
• Reported that Lovable was the best AI tool for this app type (based on their research)

Metrics / Numbers Included

• Prototype claim: AI could build an app in ~5 minutes (general intro claim)
• Effort/investment:
  • ~380 hours
  • $326 invested
• Pricing:
  • $3/month (yearly subscription equivalent)
  • $4/month (monthly subscription)
• Tooling costs:
  • $20/month Lovable
  • $25/month Framer
  • Total: $45/month
• Break-even estimate:
  • ~11 users/month at $4/month
• Iteration/security timeline mentioned:
  • ~2 weeks of break/fix cycles for live preview/dashboard reliability
  • Additional ~3 weeks for security fixes

Verdict / Recommendation

Promising concept, but not “set-and-forget.”

AI-assisted “vibe coding” can help a non-developer ship a working app—particularly the back end—but the reviewer emphasizes that front-end quality, dashboard + live preview integration stability, and security/data handling still require substantial manual design, debugging, and iteration.

If you want to build a similar creator link-in-bio product, the reviewer’s takeaway is that you can ship faster, but you should budget for UI rework and security diligence.

Unique Points Mentioned (Consolidated)

1. “Vibe coding” = building apps using AI tools only.
2. AI can generate back end well (auth, data handling).
3. AI front end is initially poor/basic; requires human UI design.
4. Built Flow Stack for creatives: one-page link-in-bio with links/photos/videos/events.
5. AI tool used: Lovable, guided with ChatGPT prompts.
6. Major prototype issues included missing pages/auth flows.
7. UI design done in Figma/Canva, with screenshot/reference guidance to the AI.
8. AI can replicate UI when given screenshots reference images.
9. Biggest challenge: dashboard live phone preview—feature additions frequently broke things.
10. Login flow bug: sign-in then immediate logout.
11. Sensitive user data stored in Supabase; security reliability is critical.
12. After more work, achieved end-to-end working + secure app.
13. Subscription pricing: $3/month yearly-equivalent, $4/month monthly.
14. Payment flow ideas considered (upfront vs free trial vs credit card timing); preference not to rely on “forgetting to cancel.”
15. Marketing plan: subreddits/Product Hunt + Meta ads (FB/IG) + TikTok organic + a “secret sauce” via wedding industry colleagues (free access → social promotion).
16. Low operating costs ($45/month) and break-even estimate (~11 users/month at $4/mo).

Speakers / Views

• Single main speaker (the reviewer/builder): covered tool choice, build process, features, costs, marketing plan, pricing, and challenges.