App
Strategy
AI Concept
2024
Strategy
AI Concept
2024
Synthesizer
AI-fueled Digital Design Platform
Working with
CPO, 4 Engineers and 2 Co-Founder
CPO, 4 Engineers and 2 Co-Founder
Synthesizer is a B2B digital design platform for luxury fashion brands, enabling teams to blend real and generative elements across ideation, production, and marketing. The platform is designed to enhance creativity, streamline workflows, and preserve brand DNA within AI-driven creation.
The Challenge
The initial vision lacked clarity around who the platform was truly for, how it differentiated from existing creative tools, and where it fit within established luxury fashion workflows.
Key unknowns included:
Key unknowns included:
-
Audience focus — designers, marketers, or production teams
-
Differentiation — how to stand apart from existing AI and creative tools
-
Core value — creative exploration vs. production efficiency vs. automation
- Collaboration gaps — siloed workflows across design, marketing, and production
The Approach
We focused on identifying where generative AI could add the most meaningful value in the creative journey without compromising brand quality or craft.
Rather than competing directly with existing tools like Adobe, the concept explored a layer above the toolchain: a system that connects ideation, brand DNA, and production logic into a single, coherent flow.
The prototype was designed to span the full creative spectrum — from concept generation to production and marketing outputs — allowing investors and early partners to react to the whole system before narrowing focus. This helped surface where appetite, traction, and differentiation were strongest.
Rather than competing directly with existing tools like Adobe, the concept explored a layer above the toolchain: a system that connects ideation, brand DNA, and production logic into a single, coherent flow.
The prototype was designed to span the full creative spectrum — from concept generation to production and marketing outputs — allowing investors and early partners to react to the whole system before narrowing focus. This helped surface where appetite, traction, and differentiation were strongest.
The Output (MVP Moodboard Interpreter)
Through investor conversations and beta client testing, we identified a clear gap: rather than using AI to automate the production of marketing assets, the real opportunity was to support the earliest stage of the creative process — ideation.
Two factors shaped this decision. First, output accuracy across generative tools was not yet reliable enough. Second, in luxury fashion, brand integrity and quality cannot be compromised. Rather than competing in a space where precision and production readiness are critical — and where other tools already perform well — we focused on building tools that help creatives translate intuition into something visible and shareable.
Two factors shaped this decision. First, output accuracy across generative tools was not yet reliable enough. Second, in luxury fashion, brand integrity and quality cannot be compromised. Rather than competing in a space where precision and production readiness are critical — and where other tools already perform well — we focused on building tools that help creatives translate intuition into something visible and shareable.
The MVP Prototype
Brand DNA Model
A key part of the work was developing a model to extract and structure brand DNA, alongside digitising brand archives to support generative ideation.
We treated past collections, runway shows, campaigns, and creative direction as living sources, translating them into structured signals the AI could work with. This allowed the system to reference a brand’s own history when generating ideas — supporting the revitalisation of past collections, the modernisation of signature elements, and the exploration of new directions without losing authorship or identity.
Rather than replacing creative judgement, the archive became a creative partner — bringing institutional memory into the earliest stages of ideation.
We treated past collections, runway shows, campaigns, and creative direction as living sources, translating them into structured signals the AI could work with. This allowed the system to reference a brand’s own history when generating ideas — supporting the revitalisation of past collections, the modernisation of signature elements, and the exploration of new directions without losing authorship or identity.
Rather than replacing creative judgement, the archive became a creative partner — bringing institutional memory into the earliest stages of ideation.