Aditya K's Work | ContraWork by Aditya K
Aditya K

Aditya K

AI Workflow specializing in Flora node-based pipelines

Submission for the ElevenCreative Challenge! I created a cinematic brand identity sequence built from a single logo, inspired by high-end cyberpunk visuals and sterile futuristic design. The concept turns abstract textures, data mapping, and reflective motion into a cohesive visual system that feels more like a premium campaign than a simple template. The goal was to create something that feels bold, minimal, and instantly usable for creators or marketers who want a full brand look from one starting point. Did this make you want to build a whole brand from one logo? ⚡ #ElevenCreative
0
116
Cover image for I gave Flora orchestration engine
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
58
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
117
Built a tool I wish existed: LogoCanvas on @Melius — drop in any logo, and it generates the entire brand world around it. Multiple Mockups, colorways, visual style. All inside one canvas. Two prompt nodes. Zero Photoshop. Came from a real client problem: "This logo looks good, but how does it live?" Now I can show 10 brand directions with one click. Everything except the logo is AI-generated. The agent even suggested better lighting setups than I would have. 🔗 Project: https://app.melius.com/projects/e24c077e-5f4e-4a1b-802a-769cec8f5c7e/canvas/24403ae6-d49a-448f-8ed0-ad2ac7b2e110 🎥 Walkthrough: 🚀
1
178
Cover image for Full Visual Identity Creation with Melius: Work in Progress
(wip) started working on a full visual identity creation using @Melius still figuring things out on how things work for final product but it will be worthit. X: https://x.com/Motion_Viz/status/2056011026420928772?s=20
1
2
205
Me: builds workflow systems that run for years Also me: stops the car to photograph some flowers Even operators need reminders: Speed is a tool. Beauty is the point. Don't forget why you're building leverage.
0
152
Cover image for nike x dystopian streetwear editorial
built
nike x dystopian streetwear editorial built entirely with AI, no studio, no crew, no budget. all build in 2 days using my Custom Flora Workflow using custom brand pipelines.
1
206
just became an official FLORA Expert on Contra. been building AI workflows for brand identity, product viz, and ad campaigns on Flora for a while now, product shots, motion, full brand systems. all production-ready output from a single pipeline. compiling my past work into a proper portfolio on Contra now, everything from e-commerce visuals to SaaS brand campaigns. if you're a brand or founder looking for AI-powered creative that actually converts, this is where I'll be posting it. more workflows dropping soon.
1
212
Cover image for Complete brand visual system for
Complete brand visual system for VERO — a premium skincare line — built entirely with Flora. Pipeline: Claude Opus + Gemini 3.1 Pro (brief) → Nano Banana 2 (generation) → Nano Banana 2 (refinement) → Magnific Precision (upscale) → Technique variations Deliverables: — Product photography (bottles, droppers, lifestyle shots) — Model + product integration (hands, face, application shots) — Cinematic mood boards — Multiple technique outputs for A/B testing Visual direction: earthy tones, botanical luxury, clean minimalism, editorial beauty. One Flora session produced 40+ production-ready assets. Traditional photoshoot equivalent: $10K-20K. This pipeline: under $30. Full brand identity system — from concept to campaign-ready visuals — in a single workflow.
1
208
Cover image for High-impact ad visuals generated using
High-impact ad visuals generated using my Flora + Magnific pipeline. Pipeline: Gemini 3.1 Pro brief → Nano Banana 2 generation → Magnific Precision upscale → Style variations Campaigns shown: — Luxury fashion (Gucci-style product shots) — Athletic wear (Nike-inspired editorial) — Bold typography layouts — E-commerce product photography Each campaign batch: 20+ variations generated in one session. Cost per visual: under $0.50. Designed for Meta, TikTok, Google display — all major ad formats. Full pipeline available as a product on my profile.
0
173
Cover image for AI-generated 3D character concepts using
AI-generated 3D character concepts using my Flora node pipeline. Pipeline: Reference images → GPT-5.2 prompt definition → Nano Banana 2 → Recraft V4 → Magnific Precision upscale Styles covered: — Sci-fi mech characters — Streetwear fashion figures — Stylized mascots — Futuristic humanoids Each character generated in under 10 minutes. Total cost per concept: under $10. Use cases: game design, animation pre-production, NFT collections, concept art exploration. Full pipeline available as a product on my profile.
0
144
Cover image for Generated a complete product photography
Generated a complete product photography shoot for AG1 using my Brand DNA node pipeline in Flora — from a single brand reference image. One image in. Four product shots out. Under $10. What makes this system different: The pipeline doesn't just generate images — it reads the brand. AG1's deep forest green, clean sans-serif typography, premium wellness positioning, and clinical-yet-warm lighting are all extracted from the reference and applied consistently across every output. A traditional product shoot like this: Agency cost: $10,000–$30,000 Team: photographer + model + stylist + retoucher + art director Timeline: 2–4 weeks The pipeline is reusable. Change the brand reference → get a completely different brand's visual DNA → same quality, same consistency. Tools: Flora, Gemini 2, Cinematic Movie Stills, XML Prompt Architecture
0
155
Cover image for Built a complete 3D streetwear
Built a complete 3D streetwear advertising campaign using Flora's node-based pipeline — zero 3D software, zero manual modeling. The system: a Master Art Director prompt architecture with 5 controllable variables (Brand, Color, Style, Environment, Render) that generates an entire campaign from a single character reference. What makes this different from typical AI image generation: → Variable-driven prompt system: swap [brand], [color], [style] and generate an entirely new campaign direction without rebuilding anything → 3D vinyl art toy aesthetic: maintained across every output — not a lucky one-off, but a repeatable style locked into the prompt architecture → Full pipeline in Flora Canvas: 15+ nodes chaining generation, style transfer, upscaling, and video in one workflow
0
147
UG — Unbreakable Vision created for @Kittl submission. Style: Noir minimalism × luxury eyewear commercial (UG — Unbreakable Vision) Mood: Dark, cinematic, precise Concept: In a world filled with noise and distraction, clarity becomes power. The character represents experience, depth, and quiet strength. (inspired by Logan) The product represents precision and vision. Together they create a cinematic narrative of focus, identity, and refinement. 🎥 Clipchamp (editing) 🎶 Eesdeekid — Century (music) 📝 Myself (script) Live on Kittl : https://app.kittl.com/project/cmmopr17p7auk0ibf539gjvwg I used Nanobanana 2 for image generation and chatgpt for image referencing. video generation using seedance 1.5 pro and veo 3.1 for final touches.
5
6
390
Cover image for trixxy | stylized character
soft storybook
trixxy | stylized character soft storybook proportions. realistic micro-detail. moody rain lighting. the breakdown: → large cranium, small nose/mouth, rounded cheeks → visible pores, slight cheek redness → damp hair clumping, wet spec highlights → fuzzy knit sweater with fiber detail → cool key light + strong rim to separate from dark bg no zbrush. no blender. no 3D software. made in @flora .
1
198
nobody has built an interaction that gives users power OVER the algorithm. so i did. reverse-engineering the algorithim patterns and accepted the makeathon challange to built something different. it's called Project Helix. instead of a feed that talks AT you, every recommendation comes with a question: "i'm showing you this because of X. do you accept, modify, or reject — and why?" you reject something → the system rewires in time. not after 30 days of behavior. immediately. the graph you see? that's the Path Matrix. it visualizes how your negotiations reshape the algorithm's decision tree. every rejection creates a new branch. every acceptance reinforces a path. you're watching your own agency become architecture. try it yourself → https://option-payer-93231299.figma.site
2
5
283
Cover image for every app you use is
every app you use is lying to you. not with words. with architecture. the like button. the swipe. the feed. they're all designed to make you say yes without thinking. i'm building something that lets you say no — and the system actually listens. here's the interaction pattern nobody's talking about: right now, algorithms work like this: system recommends → you accept or ignore that's it. there's no middle ground. no "show me why." no "i disagree because..." it's a monologue pretending to be a conversation. what if instead, every recommendation came with a negotiation frame: "i'm showing you this because of X. do you:" → accept → modify (tell me what's wrong) → reject with reason and your rejection actually rewires the system in real-time. Live Demo : https://option-payer-93231299.figma.site/
8
340
Insight Weaver — From Data to Action, Not Just Dashboards Organizations are drowning in data but starving for decisions. Insight Weaver is an open, agentic analytics fabric that turns Tableau insights into actionable, auditable narratives. Ask questions in natural language, get live visualizations, preserved reasoning, and workflow triggers. Built with extensibility at every layer — data, agents, narratives, and actions. No black boxes. Every insight is grounded, cited, and traceable. Designed for governments, enterprises, healthcare, and impact orgs. Built solo, end-to-end. Open to work, contracts, and collaborations in analytics, AI, and data storytelling.
0
159
Just made my first 1,000 dollar recreating this website for a client. While this might seem like a average product but I did the whole thing from design to coding and so far this has been the greatest experience I have had so far. Now there is no stopping, let's keep pushing.
3
4
211
Speed - TOYOTA GR USA track analysis using AI/ML and design style to predict real time analysis for a long term future prediction. This project was done for a hackathon organized by Toyota.
1
132
I built something weird, surreal, and honestly… kind of magical. It’s called Dream Gate Cinema — an app that lets you step into your dreams like movies. You enter through a glowing pillow portal. On the other side? A Dream Atlas of islands shaped by fantasy, chaos, nostalgia, and everything your mind creates at night. Tap an connected islands → it becomes a cinematic dream poster. Step inside → a Dream Cinema suggest you videos that are related to you last emotional dream. It’s not a sleep tracker. It’s an experience. A place to explore your inner world when real life feels too loud. Built entirely on @Anything Check it out: https://www.createanything.com/build/42ff6f04-55ec-43c6-8286-2f3eb8bc4d2b?view=b2e93d04-2f95-49cc-82a9-70df4b029ee7
5
171
Plasma Star Gradient Effect — Figma I developed this interface design by pushing Figma’s gradient capabilities beyond the standard tools—delivering a visually distinct result within a basic, minimalist framework. The project was created for a client seeking something unconventional, yet clean and refined.
1
135
Portfolio Design crafted over several months, exploring the biggest role of music in the emotional impact of visual storytelling. In this project, I’ve showcased that effective design is not just about compelling visuals—the integration of sound is equally vital in conveying meaning and evoking feeling. "Making you remember the feeling of connection" Discover the experience: itsadi.me (http://itsadi.me)
1
136
Cover image for How Habits, Environment, and Rhythm Shape Our Nights As a ca...
How Habits, Environment, and Rhythm Shape Our Nights As a cat owner and anime enthusiast, I’ve observed how my pet’s unique routines and the mood from anime influence my creative process and design aesthetic. By reflecting on these habits and the environments we curate, I explore how rhythm and personal interests can subtly shape our nightly rest and inspire distinctive design styles.
1
129