TLDR:

• AI stylists are moving from suggestion to decision

• Tech companies like OneOff, NewGen and SoundHound AI are redefining how we discover and buy

• Structured catalogues and a clear conversational tone are key to fully leveraging

----

In 2025, the rise of voice and chat-based AI is reshaping how fashion is discovered, styled and purchased. What was once the domain of human stylists or clunky product filters is rapidly giving way to conversational agents capable of understanding your taste, guiding your wardrobe decisions, and completing the checkout.

These aren't the generic chatbots of the last few years. Powered by LLMs,  fashion-specific datasets and commerce integrations, the latest wave of AI stylist agents don’t just respond - they listen, recommend, transact and learn.  We’ve talked a lot about personalisation in this briefing - the different approaches, the size and scale of the opportunity. This week, I wanted to take a closer look at three tech companies outside the FAANG orbit (we looked at the innovation from Google a few weeks ago) that are leaning into this shift via AI conversation and agents, both text-based and voice-enabled.

The Present: Static Filters, Slow Choices

Online fashion shopping has long been frustrating - a compromise. Sorting by price or size helps, but fashion requires nuance: tone, mood, occasion, aspiration. While algorithms have powered “you might also like” suggestions for years, the experience has lacked the warmth, intuition, and interactivity of a real stylist or in-store interaction. Until now, consumers have relied on influencers, styling apps, or trial and error.

The Future: Agentic AI with Taste

The shift we’re seeing: the next generation of AI doesn’t just answer questions - it makes decisions. Three startups are leading the way:

OneOff – The AI Stylist with Celeb Influence

OneOff blends generative AI with the cultural capital of celebrity style. Enter your favourite celebrity - say “Lewis Hamilton” - and the system returns real-time suggestions across luxury and high street platforms like MyTheresa and SSENSE, matching what you might have seen Lewis wearing on Instagram. Importantly, human creators still play a curation role, offering a blend of trust and tech that appeals to younger consumers looking for both speed and authenticity. The company earns affiliate income; its beta launched last month.

NewGen – Conversational Commerce Infrastructure

NewGen is building the backend to power “Talk to Buy” storefronts. Rather than clicking through static product pages, customers interact with AI agents that browse catalogues, suggest fits, and complete checkouts. These agents learn your preferences over time and handle everything from inspiration to payment. Their mission is to bend the internet to you - a future where interfaces are personalised and powered by LLMs, making the web dynamic and adaptive to individual users. The product has just launched and is available via API.

SoundHound AI – Voice Agents with Buying Power

Already used in the automotive and QSR sectors, SoundHound’s voice AI lets users place complex orders, handle service requests and make purchases hands-free. Their latest platform, Amelia 7.0 released, enables full transactions via speech. Imagine asking, “Do you have those white trainers in size 9?” from your car or smart home and getting a real-time response with checkout built in. With 225 patents granted or pending and 25 languages supported,  SoundHound has deep capability in the fields of advanced conversational intelligence.

What it means

The pain, friction, and compromise of online fashion shopping is longstanding - not new. What’s changed is the set of realistic, scalable possibilities. And that puts the onus on brands, retailers, and marketplaces to prepare:
Are your catalogues structured for AI agents to interpret?
Can your systems support real-time voice transactions and conversations?
Do you have a brand tone that translates well through a conversational agent?

From Discovery to Decision

Looking ahead, the next 12–18 months will be pivotal. What feels experimental today – automated styling, conversational checkout, AI-driven taste - is about to become the new baseline. I’m convinced we’re moving from manual discovery to agent-led decision-making, where curation is intelligent, instantaneous, and increasingly autonomous. Whether it takes shape through chat, search, voice, or in-app assistants, the convergence is coming. 

How do you see it unfolding?