How I use ChatGPT to run my fashion business
Yana Welander demonstrates how she built an AI-native fashion brand using ChatGPT and Codex as her technical co-founder, automating design, production, and business operations from sketches to manufacturing and e-commerce. She shares how AI unlocks previously impossible garments while highlighting remaining challenges in pattern generation and realistic image rendering.
Summary
Yana Welander is building an AI-native fashion brand that leverages artificial intelligence throughout the entire product lifecycle. The conversation explores how AI democratizes fashion design and production, removing traditional barriers to entry.
Design Process: Yana typically starts with hand-drawn sketches, then develops a detailed 'fashion prompt' that describes garment specifications including silhouette, volume, fabric behavior, construction details, movement, and even sound. She feeds these prompts into ChatGPT with image generation, iterating through different formats like product photos, Vogue-style runway shots, and influencer photos. She chose Image Gen 2 specifically because it follows visual direction and sketches more accurately than competing models, while maintaining the unique design aesthetic she needs—other models tend to homogenize designs to look like existing runway pieces.
Production and Manufacturing: For complex pieces, Yana uses Codex with computer use to generate CAD files for 3D printing, as seen with her Ruth Asawa-inspired dress featuring intricate sculptural elements that would have been previously unmanufacturable. She uses AI as a research assistant to identify manufacturing partners, drafting vendor outreach emails, and managing the pre-order process. She works in parallel with human pattern makers, treating AI and humans as competing resources—whichever delivers fastest and most accurately to her vision wins.
Technical Co-Founder: Yana built her entire e-commerce website using Codex without hiring engineers, implementing payment processing with Stripe, creating databases to track customer votes on designs, and building dashboards. She can now iterate on technical requirements using voice commands, asking Codex to find her GitHub repository and implement changes. The brainstorm of her 9-year-old building websites with ChatGPT Sites underscores how this democratizes technical capability.
Remaining Friction: Two unsolved challenges persist—getting image models to create unique, realistic renderings that align perfectly with her vision (sometimes producing overly flat or paper-like results), and generating accurate sewing patterns based on designs. For the latter, she continues running parallel processes with human pattern makers since AI hasn't solved this completely.
Key Insights
- Yana selected Image Gen 2 over competing image models specifically because it follows sketches and visual directions more accurately than alternatives, while other models tend to homogenize designs to look similar to existing runway pieces rather than unique designs
- Creating a detailed fashion prompt that specifies silhouette, fabric behavior, construction details, movement, and even sound acts as both a design specification and an AI instruction manual—comparable to how product specs guide human teams
- Yana uses Codex with computer use to operate 3D design software autonomously, generating CAD files for previously unmanufacturable garments without needing to learn the software herself
- Pattern generation remains unsolved—Yana runs humans and AI in parallel for pattern making, treating them as competing resources to determine 'who's going to get to making my thing fastest and most accurate'
- Yana built her entire e-commerce website with Stripe payments, customer voting databases, and real-time dashboards using Codex without hiring engineers, demonstrating how AI serves as a complete technical co-founder
Topics
Transcript
[0:00] Normally, I would hire a technical team, like engineers. I didn't, right? I sort of just came to Codex and [music] asked it to build the website. The idea is to use AI really anywhere in the process where that makes sense. It's literally my technical co-founder. >> There's so much you couldn't do before, practically, from a cost perspective, even from an execution perspective, that is now totally possible. And so, you can like unlock your creativity in a way that wasn't possible before. >> does unlock so much creativity. A lot of [0:31] my designs start from a hand-drawn sketch, but the other piece of this is I sort of developed the fashion prompt that's behind it.…
Full transcript available for MurmurCast members
Sign Up to AccessMore from How I AI
The biggest barrier to AI adoption is not fear of technology. It's the absence of muscle memory.
The speaker argues that the primary barrier to AI adoption is not fear but lack of muscle memory—the habitual use of AI tools. Success requires building collaborative practices through empathy, accessible training, and establishing regular workflows rather than overcoming technological anxiety.
Staying in Gmail means none of your AI work compounds
The speaker discusses how staying within Gmail limits the compounding value of AI-assisted work and writing, and demonstrates a workflow using Claude to enhance Gmail's functionality by adding formatting, links, and draft management capabilities.
Intent engineering beats prompt engineering every time
The speaker emphasizes the importance of integrating AI, specifically Claude, into daily workflows by creating reminders to seek its assistance. This approach allows for collaborative problem-solving, transforming static prompts into dynamic conversations.
Claude Code for normal people: skills, voice mode, and how to collaborate with AI
The episode discusses how AI tools like Claude can enhance productivity and collaboration in business and personal tasks by enabling users to focus on intentions rather than complex prompting. Grace Clark offers insights on her projects and how she built tailored solutions to improve communication and client relations using AI.
Running evals on your internal agents is how you keep them trustworthy over time
The evaluation of internal AI agents is crucial for maintaining their reliability. An internal eval platform allows engineers to review agent performance regularly and ensure their responses are accurate, particularly in critical areas like coding.