Aditya K
AI Workflow specializing in Flora node-based pipelines
An exploration of how a strong hero section can establish the entire mood of a digital experience.
For this project, I wanted to blend classical imagery with a more contemporary, cinematic web direction, large editorial typography, restrained UI, immersive motion, and a hero that immediately sets the visual language for everything that follows.
I designed and developed the experience while experimenting with Midjourney × Krea during the visual exploration process, then carried that direction through the interface, interactions, and final build.
A study in art direction, web design, motion, and development and how new creative tools can make the path from an early visual idea to a working experience significantly faster.
Role: Creative Direction, Web Design & Development
Exploration: Midjourney, Krea
Build: Next.js
Still exploring where this visual language can go.
0
27
Visual Identity: AI OS (WIP)
For decades, operating systems have been designed around applications. Windows.
Icons.
Menus.
Files.
But AI is changing that interaction model.
Instead of navigating software, what if we simply described an outcome?
The interface becomes conversation.
Applications become implementation details.
The computer becomes an autonomous collaborator rather than a collection of tools.
This is my current visual exploration for AI OS—an AI-native operating environment where voice, memory, and autonomous agents become the primary interface.
Still very much a work in progress, but I'm exploring how a visual identity can communicate concepts like autonomy, orchestration, trust, and intelligence without relying on the familiar operating system metaphors we've used for decades.
Curious where this evolves.
Breaking the old pattern to start a new one..
#BrandIdentity #CreativeDirection #VisualIdentity #AI #ProductDesign #MotionDesign #FutureOfWork
4
3
124
I gave Flora orchestration engine one structural prompt and it generated a complete, continuous fashion editorial sequence for the Motion Viz digital archive.
Character continuity, 65mm film emulation, dynamic weather interactions, and precise camera tracking—all managed without manual context switching between models. Base geometry was established via Gemini, photorealistic textures via ChatGPT, and temporal physics translated flawlessly by Seedance 2.0. It captured the exact alpine brutalist aesthetic I engineered, faster than standard rendering pipelines.
Fully integrated AI workflows are still scaling, i am currently testing the absolute edge cases of what this stack can do.
If you are an AI filmmaker or technical director looking to push temporal translation and multi-model generation, get involved with the technical challenges on Contra.
0
151
ongoing work: MONAD energy drink — full brand world builing using a single logo mockup
butter scotch. 110% energy. zero jitters.
what i am building:
→ vintage poster design with distressed typography
→ mockup design for the can
→ product photography (on-model + studio)
the tiger mascot with sunglasses? that was the spark. everything grew from there.
building a full world — from a single product concept to campaign-ready assets.
tools: GPT Image 2 + Seedance 2.0 + Affter Effects
timeline: 7 days day
output: total of 16 campaign assets + brand films
this is what Direction + AI brand building looks like in 2026.
3
0
199