Project overview
Atineo is a brand of MTechnology and Tocco de Amore, the personalized gift store whose website we had already redesigned. Tocco de Amore brings what it knows about putting gifts together and delivering them; we bring the whole product: the public website, the engine that decides what to suggest, the assistant, payments, the operations back office and the app. It is not a client commission: it is our own software, and everything on this page was designed, built and is maintained by us.
The challenge
A gift store competes on price or on catalog. Atineo wanted to compete on something else—getting it right—and that came with three problems an online store does not solve:
People do not forget the gift: they forget the date. The last-minute purchase is the generic one. If the system does not remember on the person’s behalf, everything else arrives late.
“AI recommendations” are usually a black box. The gift-giver needs to know why this idea and not another. A suggestion without a reason reads like an ad.
Holding a customer’s money demands trust. The savings pocket only works if the person can always see how much is there, where it came from and that they can take it out.
Goals
Saving a person should take a minute; the reminder should arrive in time and already carry three ideas within budget; each idea should explain why it fits; the money in the pocket should be visible and withdrawable; and the same platform should serve an individual, a company with its workforce and the workshops that make the gifts.
The product, screen by screen
One minute per person, including the no’s
Saving someone takes a name, a relationship, a date and two or three things they like. Also what they definitely do not: ruling things out is half of getting it right. Each preference is stored as a signal with its evidence.
The engine matches tastes against the catalog
A person’s signals are matched against the partners’ catalog. The engine proposes nothing without at least two signals and a relationship, and every idea comes with the reason it fits.
A gift pocket you can understand
Anyone who wants to can set aside an amount each month, so the gift is already paid for when the date arrives. Every movement is labeled, and the money is a liability on the books: it never funds operations.
You start from the date, not the product
Browsing starts with the occasion, which is how gift-givers think. Each occasion has its own page and tone, from the date you cannot forget to the just-because gift.
The same logic, for teams
Atineo for Business applies the engine to a workforce’s birthdays, work anniversaries and welcomes: a different gift per person, prior approval and a single monthly invoice.
Every idea comes with its reason
Twenty-one days before the date, three ideas within budget arrive. None is a gift card or a generic present, and each explains in one line why it suits that person.
What we built
Artificial intelligence, where it adds value
What sets Atineo apart is not AI: it is curation and each person’s taste profile. That is why the work is split. A language model talks with the person looking for a gift and understands what they write; what decides the suggestions is the engine, with rules that can be audited. When the person reacts—“I like it”, “not convinced”—the rules handle it without going through the model. And everything the model writes is displayed as text, never as code: it is treated the same as anything a visitor types.
Money, with no shortcuts
A platform that holds its users’ savings cannot have rounding errors. Amounts travel as whole pesos from end to end—in the database, the API and the payment gateway—, the charging rules have tests checked on both the server and the app, and each account can only read its own data: someone else’s resource responds as if it did not exist.
Technologies used
The time frames and quantities on this page—21 days’ notice, three ideas, at least two signals—are product rules, not usage results. The people, balances and gifts shown in the screenshots are samples, as the site itself states.
Have a product idea?
Tell us what you want to build and we will show you how we would do it, from the first screen to the app in the store.
This project came out of our custom software and artificial intelligence services.