# S01E03 · Transcript — Seunga Lee Full transcript of [The Useful Podcast](https://useful.ventures/podcast) Episode 3, a founder-to-founder conversation between host Gokul Bala and Seunga Lee, founder of MUST, an AI personal styling app for men. Back to the [episode page](https://useful.ventures/podcast/seunga-lee-must). This transcript has been lightly edited for readability. Filler words and false starts have been removed and the meaning has been preserved. Any errors are mine, not Seunga's. ## Spanish, Seoul, and a band career **Gokul** I'm kind of curious about your background, because even before fashion, you chose to study Spanish of all things. Why Spanish, and especially in Korea — is that a normal choice, or unusual? **Seunga** It was very niche. I'm from South Korea, born and raised in Seoul, and I lived there until about two years ago. Most students pick Chinese or Japanese as a second language, since English is mandatory in Korea. I actually studied Japanese first, but when I applied to university I thought: I speak Korean, I speak English, and if I could speak Spanish too, maybe I could broaden my world significantly. So I applied for a Spanish major and studied it hard. **Gokul** Korea has a very studious, high-pressure culture around what you study. How did your parents take it — Spanish instead of the standard doctor-or-engineer path? **Seunga** I was actually preparing for four years to get into a foreign-language high school, a kind of premium private school focused on languages, but I didn't get in. From there I kept thinking about learning a language other than English, and Spanish just matched my vibe and their culture. My mom was actually the one who encouraged me to study Spanish rather than Japanese or Chinese. **Gokul** You're also a translator of a different kind — you were in a band, you sing, you play bass. How does all of that fit together for you? **Seunga** I love singing, I love dancing, I love music, so I had to do it. I wasn't that interested in pure study like a lot of Korean students — I wanted to try something new. I was a vocalist and electric bass player in a small university band. **Gokul** Did that lead you into fashion, since you went there right after? **Seunga** Probably, but I don't think it's that direct. Even though my major was Spanish language and literature, I kept exploring — I considered switching to hotel business management, then thought about a double major in international trade, something that could connect Korea with other countries, maybe Spanish-speaking ones. I ended up joining a fashion OEM company partly because they had production facilities in Guatemala and Nicaragua, Central America, and speaking Spanish was a big plus for me there. I'm being honest — that was part of it. ## Nicaragua and burnout **Gokul** What did that fashion OEM job actually show you about how clothing gets made? I used to work as a production planner myself, so I have a sense of it — especially during COVID. **Seunga** Lots of headaches, lots of drama, lots of accidents — quality issues, client complaints. I wasn't a designer; I did client management, procurement, quality checks, color approval, and communication between factories and clients. A little bit of everything. I had to be a good multitasker and handle unexpected situations constantly. I did that from headquarters in Seoul for three years, and then they sent me to Nicaragua as a production manager for two years, right as COVID started. **Gokul** How did that change you — new country, new role, in the middle of a pandemic? **Seunga** Yes and no. I really loved working with my local staff — we spoke Spanish together, and I had to become a leader almost overnight, even though I wasn't a director back at headquarters. I focused on what I could do for them, how to motivate them. But it was tough: I'd work from around seven a.m. until sometimes eleven p.m. or one a.m., because of the time difference with Seoul. During the day I worked with local staff at the factory; in the evening, Korea time, I had calls with the sales team. **Gokul** What did COVID actually change on the ground? **Seunga** I arrived in Nicaragua right as Korea started shutting down — I flew out when Incheon Airport was maybe two-thirds closed already. On the flight there was almost no one; I could lie down across three seats. No tourists anywhere. On the production side, everything shifted to e-commerce because people couldn't go to stores, so we had to handle a lot more individual packing and a lot more manual workforce. By the time I came back to Korea after two years, I was completely burned out — physically and mentally exhausted, and I couldn't fully recover. I decided to quit and move to Canada, since I wanted to live in an English-speaking country after years in Spanish- and French-speaking ones. ## Sitting through eight hundred color consultations **Gokul** That brings us to MUST. You used your translator skills again — sitting between clients and an analyst as an interpreter at a color analysis studio. How many sessions did you sit through, and what did that do to you? **Seunga** Before I moved to Canada, I wanted to spend time with my family, so I looked for something part-time. I'm a certified colorist — I studied that after coming back from Nicaragua — but I didn't want to start a full personal color analysis practice from scratch. Since I spoke English, I found work as an English translator at a personal color analysis lab. Have you heard of personal color analysis before? It's a color theory that, as far as I know, was invented in Japan and became huge across East Asia — Japan, Korea, China. Every person has a different skin color, hair color, and iris color; combine those three and you can place someone into a category. In my app, MUST, I use nine different personal color types based on things like color temperature, brightness, clarity, and contrast. It's not about a brand or a product — it's about you. You get your best and second-best color type, which becomes a guideline: this is the color that should be near your face — hair color, eyewear, earrings, makeup. The lab had a lot of foreign customers, some flying in from Canada, Vietnam, the US, just for a session. They had English, Japanese, and Chinese translators. I was the English one, and I sat through about eight hundred sessions over roughly a year. **Gokul** What did that show you? **Seunga** Because I was the English translator, I could compare how different consultants ran their sessions — the process, the quality — depending on their knowledge and experience. I could also see the difference across nationalities in a way a Japanese or Chinese translator maybe couldn't. What I saw was that people were paying the exact same amount of money but getting very different results, depending purely on who happened to be running their session. That made me sad. I wanted to make this more accessible and, more than anything, more standardized — stable quality regardless of who's doing the consulting, and a stable amount of information the customer walks away with. **Gokul** Why men specifically? Have you just seen a lot of men who don't know how to dress, or was it more of a market opportunity? **Seunga** The very first reason is personal — my older brother, my dad, my uncles, my boyfriend. They always ask me, or their girlfriends, "babe, what should I wear?" It's a constant default question. Second, from working in fashion and at the color analysis lab, I saw that almost every fashion and beauty business focuses on female customers. Maybe that makes sense commercially, but the people who actually need this kind of service could just as easily be men — they might be shy, too busy for an in-person consultation, or uncomfortable sitting next to a female consultant while she holds color drapes up to their face and compares things. I wanted to help people save money, save energy — high efficiency, and everyone deserves to look good when they want to. ## From ChatGPT-as-translator to builder **Gokul** What made you decide this should be a piece of software, rather than a book or a social media channel or something else? **Seunga** That was honestly unexpected. I wasn't an active AI user before last December. My boyfriend told me AI was going to be huge and I should learn it. I was thinking about taking a formal course, but he suggested I just go to a meetup instead. So I went to a Lovable meetup — my very first tech meetup — last December, and then kept going to more AI and tech events. I started vibe coding in January. My friends there were learning the tools too, and they told me: start with something you already know well, your own domain expertise. So I thought — fashion, probably — and I started building MUST at the end of February. I built the product myself. **Gokul** What was that like, building software with no technical background? **Seunga** Magical. I never imagined I could build a website. Before I used it, I genuinely thought ChatGPT was a translator. Everything that's happened in the last three or four months has felt like magic to me. I went to a Lovable meetup and learned something mind-blowing, then went to a Claude community meetup and heard people talking about "Claude Code" and "open Claude" and thought — this is so fun, what is happening here. **Gokul** How have you found using AI in practice — you've said before that it doesn't remove the need for a creative director, it just changes what the job looks like. **Seunga** I don't think I'm an AI expert yet — I'm still learning, still dealing with a lot. What I've found is that I have to review everything, whether it's generating images or planning a marketing strategy. Hallucination is overwhelming sometimes. I'll believe something because "Claude said A is true," then talk to a friend who says no, that's wrong, and I go back and Claude says, "I'm sorry, I made a mistake." Every five minutes it can make a mistake. So I have to be the one who reviews, who filters — human in the loop. AI can make life easier and faster, but you still need your own criteria for what's true. ## Building MUST solo **Gokul** Walk me through how MUST actually works — the color types, the scoring, and the hard part of encoding your own taste into the system. **Seunga** I built it on Lovable at the beginning, then started using Claude for prompting to save tokens, then Cursor and v0. The whole process is simple: you take three photos, one selfie and two full-body photos. From the selfie, the AI detects skin color, hair color, and eye color, analyzes the combination, and defines your best and second-best color. From the two full-body photos, it measures your proportions — shoulder width, waist, leg length — and calculates ratios. You end up with a full report: dos and don'ts, a recommended neckline, collar shape, button style, accessory colors and patterns. I don't have a technical background, so I can't always verify whether a result is accurate, but I've improved the flow a lot from the first version to what I'm running as a pilot now. I do run into technical difficulties, especially figuring out things like Supabase on the back end. I think I need a technical advisor. **Gokul** There's usually a way around that. Even without being that technical myself, I've found that most of these platforms — Supabase included — now give out CLIs and MCPs specifically for coding agents. Once you have an account and an API key, you can hand that over to Claude or whatever agent you're using instead of clicking through the dashboard manually. Half the difficulty is just not knowing the vocabulary of what's on the page versus what you're trying to do. **Seunga** That's cool, I'll try it out. I feel like I'm connecting what I know with what I can use right now, but I clearly need to work more on it. ## Demoing on stage **Gokul** You demoed your product at a meetup recently. Demoing in person is a different animal — tell me about it. **Seunga** I love demoing in front of people — I feel alive doing it. I don't think I've enjoyed a job this much before. I've been a vocalist and bassist on a stage before, so I like being there when I'm confident about what I'm showing. The best comment I ever got was, "your demo was so fun," and it kind of shocked me. **Gokul** Why was "fun" the compliment that stuck with you? **Seunga** This was at Shopify Sunday, and I did a live demo on stage. As an audience member myself, it's usually hard to stay focused on someone else's demo if it's not your sector — you listen for ten or fifteen seconds and drift off. But fashion, dressing well, is universal. If I demo this in front of you, you think of yourself, or your guy friends, or a woman thinks of her boyfriend or brother. People could relate to it immediately, even though it's a fairly niche product for men specifically. And it worked on stage — the internet connection held, no technical problems, which is a bigger deal than people realize. **Gokul** You've hinted at this already — you're doing this alone. Product, tech, sales, marketing, fundraising, pitching to investors, even getting pulled on stage to pitch for a CTO. What does building solo actually do to you? **Seunga** I'm enjoying it a lot, but it can be overwhelming, especially around anything technical — security, privacy, making the back end more stable. That's my biggest concern and the main reason I want to find a CTO. I'm at a very early stage — I ran a pilot around Toronto Tech Week at the end of May, and it's July now, so this has all happened very recently. I can handle small technical problems myself for now, but I do need a technical advisor. What I'm focused on is what I can actually do: I know how to make content, so I'm slowly building out a go-to-market strategy through content, and building enough of a user number that I can pitch to brands or investors. ## Growing an audience from zero **Gokul** In the last six months your content also blew up — a Spanish-language account. What is that account, exactly? **Seunga** I speak Spanish in it, so people probably wonder what I'm doing and why they're following. I started making content last December — me speaking Spanish, introducing Korean culture and language to Spanish speakers. I tried teaching Korean directly, but the languages are so different it got complicated, so I started borrowing songs from Korean-to-Spanish translation content creators — the "hi, annyeonghaseyo" type lyric content — and made lip-sync videos on top of that. It blew up. **Gokul** Did you expect that? **Seunga** I thought it would be niche. There are a lot of Korean creators speaking to Korean audiences, and plenty speaking English, but I wanted something more specific. I speak Spanish, and there are people out there who want to learn about Korean culture and K-pop — thanks to idols, movies, and animation, Korean content has become a huge trend. I thought the timing was probably right, and it worked. **Gokul** A lot of founders I talk to struggle with exactly this — distribution. It seems like you built a real muscle for it before you even needed it for MUST. **Seunga** I think it's all about positioning — finding something very niche and specific, targeting a specific audience. I failed a lot of content along the way — I have something like fifteen different Instagram accounts I tried and abandoned. I wanted something I could do regardless of location or time zone, since I travel a lot. **Gokul** What actually drew you into making content in the first place? **Seunga** I opened that account while I was in Abu Dhabi. I'd moved to Canada in December 2024 and worked for Hyundai Glovis Canada — a Korean company's Canada office — for about six months. I loved my Korean colleagues, we're still close friends, but I realized I wasn't experiencing what I wanted to experience in Canada: I was reporting in Korean, writing emails in Korean, eating Korean food with Korean colleagues. My boyfriend travels a lot for business, so we traveled together, and I wanted something more flexible and remote — but it's hard to get a remote job if you're not an engineer or a designer. I was in Dubai and Abu Dhabi feeling a bit down, with nothing to really do while my boyfriend was busy. So I decided to make content. It took time — two to three months before anything really happened. My daily routine became: make one reel a day. I posted one reel a day until I hit 10,000 followers, which took about a hundred days. If someone commented, I'd reply and try to have a real conversation. I focused hard on consistency and engagement. After that it kept going in waves — up and down, up and down. I studied other influencers constantly: what kind of content blew up for them, what strategy they used, what topics about Korea were popular specifically with a South American audience. When something worked, I'd stick with that format for a while, then go research and try something new. The lip-sync format I do now is one of those. ## Why men, and the mission ahead **Gokul** If MUST works exactly the way you want it to — say five years from now — what changes for me as a guy waking up in the morning? **Seunga** You'd be more decisive, more confident about styling yourself or buying things — spending money on the right things. I want to be clear though: the point of MUST isn't that you have to be stylish all the time. I love sweatpants and hoodies too. But when it matters — a black-tie event, a business presentation, a first date, or when I have to demo MUST in front of somebody — I can pick the right clothes and styling for the occasion, because I understand myself. That's what I want to give people: an understanding of themselves. My expertise is fashion, beauty, and color analysis specifically, so that's the industry I want to help people in. **Gokul** I think that shift matters — the first time you realize you can look this good, it changes your confidence, and that ripples into everything else. **Seunga** Exactly — it's more about confidence, self-esteem, self-assurance. Everyone deserves to feel good when they want to feel good. **Gokul** Last question — I feel like nobody who tries to build something is really doing it for the money. It's usually more of a mission. What's the secret goal underneath all of this for you? **Seunga** It's gonna be gigantic. I want to be a real founder. I still feel like an explorer, honestly, because — I'll be honest — I'm an immigrant from Korea. What I have right now is an open work permit. I have to prepare a PR application, which likely means becoming an employee again, and that got me down for a couple of weeks. But I did a demo again a few days ago and found out this is my real passion. I want to keep working on MUST even if, for now, it has to stay a side project while I work another job and bring new ideas back into it. Following your passion and doing what you actually want to do — I think that's very important. **Gokul** Is there a startup pathway for permanent residency that might help? **Seunga** I've been researching it. I live in Toronto now, so I keep checking Ontario's provincial nominee programs, but they haven't announced new intake rules — I think something's expected around July or August. What I know so far is I'd need to study French and get a year of full Canadian work experience to qualify. Reality isn't making it easy, but I'm going to keep pursuing this, even as a side project for now. We'll see. **Gokul** That makes sense. Well, that's everything — thanks so much for coming on. **Seunga** Thank you so much. --- Back to the [episode page](https://useful.ventures/podcast/seunga-lee-must) · [[podcast|All episodes]]