'Meu Atelier by FashionTown' is a controlled Alpha for AI-assisted fashion product development. It analyzes visual references, preserves professional inputs, estimates preliminary fabric consumption, identifies uncertainty, suggests compatible construction options and generates traceable tech packs. This first release tests the core workflow before expanding toward sketch interpretation, technical drawings, patternmaking and grading.
What real task does your product handle with GPT-6 Astra?
Maker
GPT-6 Astra analyzes fashion references and user-provided garment data to help turn visual design intent into structured technical decisions. It identifies what is visible in the reference, distinguishes user-provided information from AI inference, flags missing or ambiguous construction details, suggests compatible options when needed, and generates a traceable preliminary tech pack for professional review.
Report
Maker
📌
Hi Product Hunt! I’m Vanessa, founder of FashionTown.
Meu Atelier is the first controlled Alpha of a larger AI-assisted fashion product development system we’re building.
The long-term goal is to help fashion professionals turn sketches, visual references and design ideas into structured technical outputs — including technical drawings, tech packs and, later, patternmaking and grading workflows.
For this first test, we deliberately narrowed the scope to three blouse constructions. We want to understand how the AI behaves when it needs to interpret visual references, distinguish observed information from inference, ask for missing decisions, estimate preliminary fabric consumption, and suggest compatible construction alternatives without taking control away from the professional.
This Alpha can currently:
analyze a controlled fashion reference;
preserve user-provided materials and measurements;
estimate preliminary fabric consumption;
account for construction choices such as gathers and bias finishing;
suggest compatible back constructions when they are not visible;
generate a structured, traceable tech pack.
The goal of this launch is not to present the final system. It is to test the core interaction: can AI help translate a fashion idea into technical decisions while keeping the professional in control?
Our principle is simple:
AI proposes. Professionals validate. Users decide.
We’re using this Alpha to learn where the model performs well, where ambiguity needs human clarification, and what should be improved before expanding toward sketch interpretation, technical drawing generation and patternmaking.