The Profile
Jen Rhi
She looked at a fashion campaign and saw a data pipeline. Then she built the machine to run it.
Jen Rhi, drawn for Forbes Korea's rising-AI-leaders list. The smile of someone who found the bottleneck.
She looked at a fashion campaign and saw a data pipeline. Then she built the machine to run it.
Somewhere in a luxury brand's marketing department, a campaign that used to require a studio, a lighting crew, a stylist, a booked model and three weeks of calendar is now a file that renders overnight. The garment is real. The fabric drapes correctly. The model looks like a person. Nobody flew anywhere. That quiet substitution is the business Jen Rhi has built.
Rhi is the founder and CEO of NXN Labs, an AI company split between San Francisco and Seoul that makes photorealistic on-model imagery, video and virtual try-ons without a physical photoshoot. The pitch is not that the pictures are cheaper. It is that the whole factory of fashion imagery can become software - generated, tested and shipped through one pipeline.
She did not arrive here from the art side. Rhi trained as a computer scientist at Stanford, advised companies at McKinsey, and sat on the investor side of the table at SoftBank Ventures. She spent years studying startups before deciding the more interesting seat was the founder's chair. NXN Labs is what she got up and built.
Creative wasn't the constraint; production bandwidth was.- Jen Rhi, on why NXN Labs exists
Sources: NXN Labs, PR Newswire, Naver D2SF, Forbes Korea.
Ask most people why fashion e-commerce is slow and they will point at creativity - the hunt for the right concept, the right mood, the right look. Rhi points somewhere less glamorous. The concept is cheap. What costs money and burns weeks is producing the image: booking the shoot, dressing the model, lighting the set, retouching the file, then doing it again for every colorway, every market, every A/B test.
That is the bottleneck NXN Labs attacks. Its models generate the model shot, the lookbook, the campaign frame and the virtual try-on from the same engine. A brand can spin up variations, test which visuals actually convert, and deploy the winners - without a single call sheet.
The hard part is that clothes and bodies are not separate problems. Fabric folds because a shoulder is under it. NXN Labs models the human body and the garment at the same time, which is how it keeps a brand's cut, texture and identity intact instead of producing generic AI mannequins. Two global patents sit under that method.
Arden AI is the agentic operating system - generate assets, review, approve, track performance.
NXN Studio is the managed arm - PDP shots, campaigns, films, custom virtual models, post.
Virtual models come curated or custom, and license-free.
One pipeline runs sample → lookbook → campaign → try-on.
Studies computer science. A classmate cohort that treats software as a default tool, not a specialty.
Learns how big companies actually decide - and where their operations quietly break.
Moves to venture capital, judging other founders' pitches. Absorbs the pattern of what wins.
Founds NXN Labs with CTO Lena Hong, a fellow Stanford grad and former Microsoft Copilot researcher.
Naver D2SF leads the seed round, with KB Investment and Smilegate - just two months in.
Selected for Korea's Deep Tech TIPS program. Recognized as an NVIDIA partner.
Commercial launch. Signs KREAM and Shinsegae International. Exhibits at VivaTech in Paris.
NRF Top 50 and FashionUnited Top 5. European luxury clients. Off to Shoptalk in Las Vegas.
We want to become the next Adobe of commercial image generation - a place where anyone can freely make images and video.- Jen Rhi, on the long game
Plenty of startups sell a text-to-image button. Rhi is after something structural: the layer brands run their entire visual production on. She describes the goal as generating, testing and distributing the visuals that actually sell - at scale, as a global standard. Fashion is the beachhead. Beauty is the stated next stop.
It is a big claim from a company that is barely three years old. But the early evidence is unusually stacked in her favor: institutional money before the paint was dry, patents on the core method, government deep-tech backing, and enterprise logos in a market - luxury - that is famously allergic to letting anyone near its image.
That last part is the tell. When a European heritage house lets an AI render its campaign, it is not buying novelty. It is buying bandwidth it could not otherwise afford. Which is exactly the sentence Rhi has been repeating since day one.
Her Korean name is Lee Jae-won. "Jen" is the handle she carries across borders.
NXN Labs never fully sleeps - the company keeps time on two coasts at once.
Her CTO, Lena Hong, studied cognitive science at Stanford and researched AI at Microsoft's Copilot team.
The consumer product is named Arden AI; the studio arm is NXN Studio.
The whole pitch reduces to one stubborn sentence: the photoshoot, not the idea, is the costly part.