ADIFI

• 

  home page

• 

  editing

• 

  editing

Inspiration

As a 15-year-old JEE aspirant, my life is built around study schedules, problem sets, and pressure. Everyone around me expects academic excellence, not creative expression. I’ve often heard:

“Coding won’t take you anywhere right now.” “Why waste time on side projects?”

But I’ve always known that I wasn’t built just to follow a syllabus. I wanted to build something that felt like mine.

Late one night, drained after coaching, I opened Bolt.New. No installs, no distractions — just an open canvas and a crazy idea:

What if photo editing could be lightning-fast, powerful like Photoshop, simple like Canva, and all in the browser?

That’s how ADIFI was born. Not just a tool, but my rebellion against rote learning and my belief that I can create something real — even when no one believes in me.

What it does

Adifi(AKA photo studio pro) is a lightning-fast, privacy-first photo editor that works completely in-browser. It combines the precision of Photoshop with the simplicity of Canva, without requiring any downloads or uploads.

Features:

Multi-layer editing (images, text, effects)

Real-time filters (brightness, contrast, blur, vignette, etc.)

AI-powered background removal (using BodyPix)

Instant exports (with zero server processing)

Fully responsive UI with keyboard shortcuts and live previews

How we built it

Built entirely in Bolt.New using vibe coding — described my vision and let AI scaffold the architecture.

Used React + TypeScript with Tailwind CSS for structure and styling.

Built a custom Canvas engine to handle layers, effects, and pixel manipulations.

Integrated TensorFlow.js and BodyPix for client-side AI features like background removal.

Added undo/redo functionality, image state tracking, and real-time performance stats.

Challenges we ran into

Balancing studies and coding: Working on ADIFI late at night, after long coaching days, was emotionally and mentally draining.

Performance bottlenecks: Large image processing would slow the browser — fixed by optimizing canvas redraws and memory usage.

AI unpredictability: Bolt's AI sometimes overwrote working code; I had to constantly guide and correct.

Layer positioning bugs: Aligning text and image layers across devices took hours of trial and error.

Emotional doubt: Hearing people around me mock the idea of "coding during exams" was hard. But I kept going.

Accomplishments that we're proud of

Built a fully functional photo editor using only browser tools and AI assistance.

Learned to handle complex canvas rendering, image filters, and AI integration — solo.

Shipped something real despite intense academic pressure.

Proved to myself (and hopefully others) that I’m more than just a student.

What we learned

Deep understanding of the Canvas API and image processing fundamentals.

How to work with AI models in-browser (BodyPix via TensorFlow.js).

How to vibe code efficiently using Bolt.New: starting from prompts, refining, and debugging.

The importance of resilience and consistency when no one else believes in you.

What's next for ADIFI

Add object removal, AI filters, and QR-based editing tools.

Launch as a progressive web app (PWA) with offline support.

Open-source it for students, creators, and indie designers.

Build a lightweight mobile-first version for quick edits on the go.

Keep proving that even as a student, I can ship real things that people use.

Built With

• bodypix
• bolt.new
• css
• htmlcanvasapi
• lucide
• react
• supabase
• tailwind
• tensorflow.js
• typescript
• vite