DEV Community

Nilesh Jethwa
Nilesh Jethwa

Posted on

How I Built Vizbull: Turning Photos Into AI-Generated Art

Like most of us, I’ve spent countless hours tinkering with generative AI—experimenting with prompts, trying new models, and wondering: could this be turned into something useful, fun, and beautiful?

That itch eventually led me to build Vizbull
— an AI-powered image transformation tool that lets anyone turn their photos into line art, cartoons, watercolors, or even custom coloring pages.

It started as a weekend experiment. Today, it’s a growing SaaS project with paying users, real workflows, and integrations on the roadmap. In this post, I’ll share:

✅ What Vizbull does and why I built it

⚙️ The tech stack behind the scenes

🚀 How I’m approaching monetization & growth

👀 What’s next

🖌️ What Vizbull Does

At its core, Vizbull is simple:

Upload any photo (portrait, pet, landscape, product)

Choose a style (cartoon, Tintin-style, anime, watercolor, pop-art, etc.)

Get a transformed image in seconds

But the magic isn’t just in the styles—it’s in creative use cases:

Artists use it to generate line art references for sketching

Parents turn family portraits into coloring pages for kids

Creators make cartoon profile pictures for TikTok/Instagram

Businesses design branded character art without hiring a studio

⚙️ The Tech Behind Vizbull

Since Dev.to readers love the build side, here’s a quick peek under the hood:

Frontend: Vanilla JS + jQuery with Bootstrap for fast iteration (yes, I kept it simple intentionally)

Backend: Serverless AWS Lambda (Node.js)

Image Processing: OpenAI Image APIs (edit/variation) + custom style prompts + some post-processing scripts

Storage & Delivery: S3 for images, SES for emailing results

Payments: Stripe micro-credits system (buy credits → use them per transformation)

One design principle: low friction. No complex accounts needed to test. Users can upload, try styles, and only pay when they want downloads in full resolution.

💸 Monetization & Growth

Instead of subscriptions (which can scare away casual users), Vizbull runs on a credit-based model:
(NOTE: I am adding more Photo utilities so will change to small monthly subscription)

Basic: $9 (X credits)

This matches usage patterns better—some people only want 2–3 transformations, while others batch process 50.

For growth, I’m experimenting with:

🚀 Product Hunt launch (coming soon)

🎥 Short video marketing (TikTok & IG reels showing before/after transformations)

📰 Dev-friendly writeups like this one 😅

👀 What’s Next

Here’s what’s on the roadmap:

Batch uploads (drag & drop 10+ photos at once)

Custom prompt blending (combine your style notes with Vizbull presets)

Collage & GIF maker (before/after sliders, meme-style exports)

Integration with InfoCaptor AI (my other project—summarizer + knowledge graphs for YouTube transcripts) so users can connect visual creativity with research workflows.

Check out:

InfoCaptor AI
→ YouTube summaries + knowledge graphs.
https://my.infocaptor.com

Vizbull
→ AI photo-to-art generator.
https://vizbull.com

Wireframes.org
→ AI wireframe prototyper.
https://wireframes.org

Top comments (0)