About Tre-Up

Tre-Up is an iOS app in development: a behavioral decision system for the moment your focus, energy or direction starts to change. People usually know what to do. The hard part is noticing the moment a day starts to slip, and deciding what to do next. Tre-Up is designed around that earlier moment.

It follows one chain: signals, a shift, Pursue, a next move.

Most days don't fall apart all at once. Focus usually changes gradually: you get distracted more often, simple decisions take longer, and the thing you meant to do gets pushed back without you noticing exactly when it happened. Most tools react after you already feel off. Tre-Up looks for the earlier moment.

One weekday shows it:

Between 13:15 and 13:22 nothing dramatic happens. That is exactly the problem. The day is illustrative: it shows what Tre-Up is designed for, not a record of the app in use.

Pursue is the AI inside Tre-Up, not a separate product. It is designed to notice when your focus, energy or screen rhythm shifts, and to turn the context in front of you into one clear next move.

When it notices a shift, it speaks once: one line on what changed and why it matters now, and one small step. When it does not know enough, it asks one question. When nothing needs to change, it stays quiet; it is designed to stay quiet more often than it speaks. You decide whether to act.

Tre-Up is not a task manager, a habit tracker or a general-purpose AI chatbot. Those hold what you planned, or answer what you ask. Tre-Up focuses on behavioral signals and decision timing: the moment a plan starts to slip, and one move that fits it.

Tre-Up has not been released. It is being built for iOS first. Invitations to the early version go out in small groups; you can request one on the home page.

We make no claim yet that Tre-Up works, and Tre-Up has not yet published results of its own. In its first test with real people, the main question the next day, for invited participants, is: did you take the step? The Pursue Signal Preview is a small study of how Pursue reads a moment, open to anyone; you take part with a code instead of your name.

Tre-Up is built in Istanbul, Türkiye, by its founder, Yiğit Ali Kurt. It is not yet incorporated: the privacy policy names the founder as the person responsible for your data, and will name the company once there is one.

To reach Tre-Up, write to company@tre-up.com. Tre-Up is also on Instagram and X.

Tre-Up's design rests on published research, summarised in Research Notes: 12 notes in 7 topics that answer everyday questions about attention, interruptions, timing and trusting an AI with your day. They draw on 40 studies by authors at 50 institutions in 10 countries, published 1966–2026, and 14 primary legal and technical documents. Every claim links to its source.

The notes are written by Pursue Lab, Tre-Up's research process. AI research assistants do most of the searching, reading and first drafting, a separate review step checks each summary against its source, and the founder decides what is published. Every study citation is checked against its DOI or arXiv record, or the publisher's own page or a library catalogue record when neither exists. Summaries are in our own words and link to the originals (how the notes are made).

The notes describe research and design intent, not results from the app, and they are not medical or legal advice. The institutions are named to show where the research was done; none is affiliated with Tre-Up or endorses it.

Tre-Up does not sell your data: not personal, not anonymous, not pooled. Inside Tre-Up, everyone gets a Pursue of their own; what it learns about your days works for you, and nobody else. You can delete your data whenever you want.

"Anonymous" is weaker than it sounds: researchers found 99.98% of Americans could be re-identified in an "anonymous" dataset from 15 details like age, gender and postcode (Rocher, Hendrickx & de Montjoye, Nature Communications, 2019).

The website collects very little: an email address if you join early access, coded check-in answers if you take part in the Signal Preview, and visit statistics from Google Analytics, which does not receive your name or email. The privacy policy, in English and Turkish, covers the website, its cookies, the early access list and the Signal Preview; the app will have its own policy before launch.