Inspiration

Tattoo ideas often start as a feeling, memory, symbol, or rough concept, but turning that into something visual can be difficult.

I built ATRUM to make that process easier.

ATRUM is an AI-powered tattoo design app that helps users turn written ideas and meaningful photos into personalized tattoo concepts in seconds.

What it does

Users can:

  • Describe a tattoo idea
  • Choose from 20+ tattoo styles
  • Generate a custom tattoo design with AI
  • Use a reference photo, such as a pet, portrait, or meaningful memory
  • Save, download, share, and create variations of designs

The core experience is simple: idea → style → generate → refine

How I built it

ATRUM is a native Android app built with Kotlin and Jetpack Compose.

The backend handles AI generation, usage limits, and secure communication with Replicate. Tattoo-specific prompt rules are used for each style to produce cleaner, more tattoo-ready results.

For Photo to Tattoo, the reference image defines the subject while the selected style defines how that subject is visually transformed.

RevenueCat manages premium access and subscription entitlement state.

Challenges & learnings

The biggest challenge was balancing creativity with consistency.

If the AI follows a photo too closely, the result looks like a traced image. If the style is too aggressive, the subject can lose its identity.

The key rule became:

Identity first, style second.

I also learned that a good AI product needs more than a strong model. UX, error handling, analytics, prompt design, and monetization all need to work together.

What's next

ATRUM is evolving beyond a simple AI tattoo generator.

Next features include:

  • Placement-aware tattoo generation
  • Body preview
  • AR tattoo try-on
  • Smarter design variations

The long-term goal is simple:

Help people explore their tattoo before making it permanent.

Built With

Share this project:

Updates

Submission history