This blog post provides a summarized version of our conversation with Chang Liu, featuring key highlights with slight alterations for readability. To listen to the full discussion, check out our podcast on Spotify and Amazon Music.

What if online shopping could think like you do?
Not in filters or SKUs, but in moods, context, and moments — structured and chic for a white party, or cozy but not try-hard for Thanksgiving.

In this episode of She Builds with AI: Women Shaping the Future, I sat down with Chang Liu, Co-Founder and CEO of Plush, to explore how she turned her personal frustration with endless scrolling into an AI-powered fashion discovery platform that promises less searching — and more delight.

From Finance to Fashion Tech

Chang’s career path traces a rare but revealing arc: from finance to data science to product — a trajectory that now feels inevitable in hindsight.

As an analyst, she learned to read business models and understand what makes them last.
As a data scientist, she learned to trust measurable signals over opinions.
As a product manager, she learned to ship minimal, lovable products that resonate.

That blend of analytical rigor and user empathy became the backbone of Plush — a system designed to make online shopping intuitive, expressive, and human again.

Shopping, the Way People Actually Think

Plush begins with a simple but deeply human insight: we don’t shop like databases.

Traditional filters break down when faced with the way people actually describe what they want — “something elevated but relaxed,” “romantic without being frilly,” or “a structured dress that travels well.”

Plush’s core bet is that modern AI can finally bridge that gap — translating the way people imagine outfits into clean, shoppable results that feel right.

The Technology Behind Intuitive Discovery

Three advances make this vision possible today:

1. Multimodal models that can see fashion — understanding fabric, silhouette, and finish beyond what text alone can describe.
2. Instruction-tuned language models that interpret subjective prompts and turn them into structured parameters the system can act on.
3. Stronger embeddings that map queries and brands into the same vector space — matching “modernist wedding in a Tuscan villa” to outfits with the right tone, drape, and aesthetic.

And beneath it all sits an unglamorous but vital foundation: clean, structured data. Without it, personalization stalls; with it, discovery accelerates.

Building with Evidence and Empathy

The lessons Chang carried from her past roles show up everywhere in Plush’s design philosophy:

• From investing: moats matter — in data, unit economics, and brand differentiation.
• From finance: model cash flows realistically, so the business can breathe.
• From data science: let numbers guide decisions; instrument everything.
• From product: build minimal lovable products, not just MVPs.

That mindset led Plush to rebuild its hybrid search engine from scratch — tripling retention by prioritizing accuracy, clarity, and delight over volume.

From Transactional Search to Contextual Curation

The future Plush envisions is one of contextual discovery, not transactional search.
It’s about curating a tight edit of ten to twenty results — each explained through evidence: trend cues, fabric quality, brand identity, or styling context.

This is curation with confidence — where AI acts not as a replacement for taste, but as a thoughtful stylist that reduces noise while preserving the joy of browsing.

Real Moments, Real People

The proof is in the queries.
Users come to Plush with deeply personal needs:
“Upstage the bride, but don’t.”
“Black-tie comfort for postpartum.”
“Jeans with exact in-seam, natural fibers only.”

Because Plush understands both language and style, it can meet users where they are — honoring preference, context, and even vulnerability.

Founders in Sync

Plush’s co-founders, Chang Liu and Ying Zhu, met on Y Combinator’s co-founder matching platform — and instantly clicked. Their shared background in tech and creative work created a rare synergy: data-driven precision meets aesthetic intuition. Together, they’re building a company where technical excellence and emotional intelligence coexist by design.

A Future Built on Taste and Trust

As AI continues to evolve, Plush’s north star remains steady: to translate human taste into precise, confident choices — while keeping discovery warm, playful, and personal.

For Chang Liu, the joy of fashion lies in the why behind what we love. Plush is here to make that “why” visible, searchable, and beautifully human.

---

🎧 Listen to the Full Episode

This conversation is full of insights on fashion, AI, powerful women, and creativity.
🎧 Listen now on Spotify and Amazon Music or other major podcast platforms and share it with someone who needs that spark of inspiration and vision.

✨ Stay Connected with Chang & Plush

🔗 LinkedIn Chang: https://www.linkedin.com/in/changliu/
🔗 Instagram Chang: https://www.instagram.com/misschangeychang/
🔗 Plush Website: https://www.plush.shop/
🔗 LinkedIn Plush: https://www.linkedin.com/company/plushshopping/
🔗 Instagram Plush: https://www.instagram.com/_plushshopping/

---

She Builds with AI  is a podcast that spotlights women who are shaping the future with AI and Emerging Technology all over the world.