You already own the most valuable thing in shopping. You just can't reach it.

@fridayresearch_
INGLÊShá 1 dia · 30/07/2026
2.0M
1.7K
1.6K
1.7K
40

TL;DR

FRIDAY is a privacy-focused AI shopping assistant that builds a local Taste Vault on your device to provide personalized, ad-free product recommendations across the web.

Your taste is scattered across forty companies that each own a piece and answer to advertisers. FRIDAY is building the version you own — a personal shopping intelligence that lives on your device and works for you.

FRIDAY - Personal Intelligence for Shopping - inline image
FRIDAY - Personal Intelligence for Shopping - inline image

The best recommendation you'll get today was chosen by an auction

**

**

FRIDAY - Personal Intelligence for Shopping - inline image

FRIDAY is a personal shopping agent that answers to you

FRIDAY is a privacy-first personal shopping assistant. You ask in plain words — \"minimal white trainers under £120"\ — and it returns a short, ranked shortlist that fits you, and tells you why each pick made the cut.

The defining idea: your taste — brands, styles, budgets, your visual "style DNA" — is computed and stored on your own device. There are no promoted products and no paid placements, and there never will be. A recommendation should answer to the person receiving it — not to whoever paid the most.

How it actually works — your taste never leaves your device

On the device, a Memory Engine of roughly 28 modules builds your shopping brain in your browser and keeps it there. It learns which brands you return to, the price you anchor on, the shapes and materials you reach for, and it lets old signals fade so the picture stays current.

On the server sits only the shared product catalogue, the AI, and end-to-end-encrypted blobs the server cannot read. Your identity is a key you hold — a cryptographic fingerprint — not an email account a company owns. Products are matched with both text and image AI, then re-ranked against your on-device taste: the intelligence comes to your data, not the other way round.

FRIDAY - Personal Intelligence for Shopping - inline image

Your taste builds itself, quietly, as you shop

You don't fill in a survey and hope. A Chrome extension (live in the Chrome Web Store) builds your taste on-device as you browse the shops you already use — if it caught your eye, FRIDAY remembers.

The model is living, not a signup snapshot: it updates from how you actually shop, so the shortlist gets sharper the more you use it. Nothing about that browsing is uploaded in readable form — the learning happens on your side of the glass.

FRIDAY - Personal Intelligence for Shopping - inline image

It recommends. You decide. You check out where you always have

This is the line we will not blur: FRIDAY recommends — it does not buy for you. You check out on the retailer's own site, under their security, exactly as you do today.

FRIDAY earns a commission only when you actually buy — so the business grows in direct proportion to how useful the recommendation was. No promoted slots to sell, no incentive to nudge you toward whoever paid. Incentives finally point the same way: yours.

FRIDAY - Personal Intelligence for Shopping - inline image

For an agent to be genuinely useful it has to know what you own, what you returned, what you're saving for, what you'd never wear. Today that context is split across forty silos — email receipts, browser history, retailer accounts, wishlists, returns — and no single company owns it. Whoever lets you own and carry it holds the layer everything else plugs into.

FRIDAY - Personal Intelligence for Shopping - inline image

The obvious question: why won't Amazon or Google just do it? Because their business structurally can't. Amazon optimises inside Amazon. Ad-driven platforms can't promise your taste stays on your device without cannibalising the revenue that funds them.

That's the opening. A taste you own and carry across stores, agents and devices — the vision we call the Taste Vault — is something the incumbents are the least able to build, because it means giving up the very data lock-in their models depend on. *(That portability is our roadmap; what's shipped today is the on-device, encrypted foundation it's built on.)\*

FRIDAY - Personal Intelligence for Shopping - inline image

This isn't a concept. It's live, and it's working

FRIDAY is in open early access with real, compounding traction: 14,342 live users, 10,574 personal taste graphs built, $1,023,322 in GMV and 2,200,000+ products recommended — at 21% click-through per search and 43% weekly retention. Live affiliate rails already reach 50,000+ ecommerce partners, and over 100,000 real products from known retailers are indexed and searchable (approximate, and growing).

Those retention and click-through numbers are the ones that matter: people don't just try it, they come back. The shopping internet was built to answer to advertisers. We're building the one that answers to you — and enough people already feel the difference to keep returning.

Meet FRIDAY

Ready to see it? Meet FRIDAY → heyfriday.app/start

More on how it thinks — www.friday.xyz

What is FRIDAY

FRIDAY is a personal AI shopping agent that learns your taste and works for you, not advertisers.

It learns your taste on your own device, keeps it encrypted, and narrows thousands of options down to the few that genuinely fit your taste, size and budget. It recommends only — no promoted products, no paid placements, and there never will be.

It lives on the open web, ad-free, in open early access. Come see what it feels like when shopping answers to you — heyfriday.app.

Guardar com um clique

Faça leitura aprofundada de artigos virais com IA no YouMind

Guarde a fonte, faça perguntas específicas, resuma o argumento e transforme um artigo viral em notas reutilizáveis num único espaço de trabalho com IA.

Explorar o YouMind
Para criadores

Transforme o seu Markdown num artigo 𝕏 impecável

Quando publica os seus próprios textos longos, formatar imagens, tabelas e blocos de código para o 𝕏 é uma dor de cabeça. O YouMind transforma um rascunho completo em Markdown num artigo 𝕏 impecável e pronto a publicar.

Experimente Markdown para 𝕏

Mais padrões para decifrar

Artigos virais recentes

Explorar mais artigos virais