Abhinav Awasthi
Building AI-driven UI/UX, code, and design solution
New to Contra
Abhinav is ready for their next project!
When I was little, I used to sit on a bench in the park near my house and imagine I was somewhere else entirely.
Underwater, breathing with the fish. In a city made of shadows. Standing in a temple made of light.
I think every kid does that. You sit still on the outside, but inside you're visiting five worlds at once.
That idea stuck with me. What if you could actually see those worlds? What if one ordinary photograph could crack open into five impossible realities — and you could watch a girl live inside each one?
That became "One Photo. Five Worlds."
I started with the most mundane image I could think of: a bored girl sitting on a park bench on a gray Tuesday afternoon. Then I used Melius to transform that single moment into five cinematic universes:
🌊 The Deep — An underwater kingdom where her hair floats weightlessly and jellyfish drift past her fingertips. She reaches out in wonder.
🔥 The Remains — A post-apocalyptic wasteland where she stands on the rusted bench, scanning the horizon. Everything ended. She did not.
✨ The Glow — An enchanted bioluminescent forest where she kneels in the moss, laughing as fireflies land on her shoulders.
🎬 The Shadow — A 1940s film noir city in black and white, with one amber neon sign as the only color. She looks over her shoulder like a tiny detective.
⛅ The Divine — A mythological temple bathed in god-rays, where she stands with her face tilted toward the heavens. She was never ordinary. The world just forgot.
In every world, her outfit transforms. The bench transforms. The lighting physically touches her skin. She's not placed in front of these worlds. She's inside them.
The only constant across all five realities is her face and a tiny silver star necklace.
Feedback on Melius:
Melius turned what would have been weeks of multi-tool work into one afternoon on one canvas — the visual pipeline from GPT Image 2 to Seedance 2.0 made cinematic world-building feel effortless.
Project link: https://app.melius.com/projects/41e31fc2-17c5-42ea-9393-7ca9e6a1302e/canvas/b283bda7-a6bb-4bf6-bcf4-09abfedef213
4
6
124
THE DICTIONARY OF UNNAMED FEELINGS.
1
12
Created a Web UI .
0
26
Your logo shouldn’t just sit there — it should respond.
I built Reactive Voice UI Animator to turn a simple 2D logo into a fully interactive voice interface.
It Flows through 4 states:
Idle → Wake → Processing → Action
From calm Glassmorphism to audio-reactive motion — just like a real voice assistant.
💡 The twist?
Each state is not generated separately — it evolves from the previous one.
That means perfect consistency, reflections, lighting… everything feels real.
🎬 Then video interpolation brings it all to life as smooth, production-ready animation.
⚡ Why it matters
No motion designer needed
Export as .mp4 / .lottie
Drop directly into React or Flutter
This isn’t AI art.
It’s a design system powered by AI.
🔗 Workflow: https://www.morphic.com/en/workflows/019dc5d1-2fe0-738d-9225-a69ac9bff877/reactive-voice-ui-animator
#ContraChallenge #AIWorkflows #UIDesign #MotionDesign #VoiceUI
3
7
344