ÆTHER: AI-Powered Brand DNA Engine by Talia PrzybyszewskiÆTHER: AI-Powered Brand DNA Engine by Talia Przybyszewski

ÆTHER: AI-Powered Brand DNA Engine

Talia Przybyszewski

Talia Przybyszewski

Overview

ÆTHER was built as part of a Contra creative workflow challenge sponsored by Melius. The brief: build something useful, inventive, and visually strong using Melius' AI workflow canvas.
Instead of making a single content generator, I built a Brand DNA Engine — a system that takes a vague startup concept and transforms it into a cohesive, launch-ready luxury brand ecosystem.
It focused on solving a common problem in early-stage product building: founders often have ideas, moodboards, or fragmented direction, but no cohesive brand system. ÆTHER turned that into a structured creative workflow.

What Brand DNA Engine Does

The system takes inputs like:
Startup idea or concept
Screenshots and references
Competitors
Mood, tone, and vibe keywords
Audience intent
Visual direction
Then orchestrates outputs across the full brand spectrum:
Brand positioning and emotional tone mapping
Colour direction and typography suggestions
UI direction and visual identity cues
Landing page concepts
Campaign visual ideas
Messaging systems
Luxury-style creative outputs
The core idea: turn a vague startup idea into a launch-ready design ecosystem.

How the AI Workflow Works

The system was built around connected AI generation nodes inside Melius, structured as a multi-stage pipeline:
1. Input Layer — User provides the idea, startup concept, vibe, references, and competitors.
2. Analysis Layer — AI interprets positioning, emotional tone, audience signals, and visual style patterns.
3. Brand DNA Extraction — Core identity traits generated: premium, playful, minimal, editorial, luxury, etc.
4. Branching Workflow — Separate creative branches fire in parallel: UI direction, typography, colour systems, campaign concepts, messaging tone, and visual outputs.
5. Human Refinement Loop — Outputs can be iterated and improved manually at each stage.
The key strength was AI orchestration + structured refinement, not one-shot prompting. Each branch could be tuned independently while maintaining coherence across the full brand system.

Tech Stack

This project was different from my usual Next.js builds. No traditional frontend product — it was AI systems design and workflow architecture:
Melius — AI workflow canvas and orchestration
Prompt chains — multi-stage generation loops
Visual branching systems — parallel creative output paths
Human refinement steps — manual iteration at each node

Why It Matters

Most AI brand tools generate isolated assets. ÆTHER generates systems — interconnected identity layers that hold together as a cohesive whole. It's the difference between getting a logo and getting a brand.
Like this project

Posted May 25, 2026

An AI-native brand generation engine that transformed vague startup ideas into cohesive identity systems, UI direction, and campaign-ready creative outputs through structured workflow orchestration.