Know yourself.
Grow together.
An app for singles, people who've just met someone, and couples. A few minutes a day to understand yourself, and the people you love, a little better.
We'll never give you a compatibility score. Nobody can honestly predict that yet.
Who it's for
Single
Get clear on what you actually want, before you meet them.
Just met someone
Keep the early days, and look back on how it began.
Together
See a little more of each other, for as long as you're together.
What we promise
Private by default
What you write is yours. A partner sees only what you choose to share.
No scores
No ratings, no percentages, no comparing you with anyone.
Honest about AI
An AI reads only your own answers, to write your nightly question and what your spots mean, and we say so plainly. Our guide, Megan, is a written character, not an AI.
We're in beta, and we're keeping the details close until launch. Beta testers see everything first.
How your data is used
Everything you answer builds your own record: your brain map, your journal, your monthly recap. It's there so you can see your own journey, and it's yours to export or delete at any time.
If you're in a couple, the reports, things to say, and songs work from what each of you chooses to share. Your partner never sees anything you haven't shared.
Questionnaire and check-in answers are analysed together to learn what good matches look like and to build BrainMatch's matching. We call this matching data. It's filed under a random id, never your name or email, alongside your birth year, whether you're single or in a couple, when your relationship started, how you met (if you meet someone here). That makes it pseudonymised, not anonymous: the id links your answers over time and to your partner's, so it's still treated as your personal data. It never includes your writing, your email, or your gender and the gender you're looking for. “Help build matching” controls whether yours is used to build matching features.
An AI reads your own written answers to write two things for you: tonight's question, which follows what you've been writing, and a few sentences on what your own words lit each spot on your map with. What's sent is your written answers to the daily question and the Monday Mix questions: only yours, never a partner's, a buddy's or a friend's, and never your name or email. The model is run by Anthropic, under a data-processing agreement, and your answers are not used to train its models. What comes back is shown to you alone, kept with your account, in your export and deleted with it. It's a switch of your own (“An AI reads your answers”, at Join and in Privacy): off, nothing of yours is sent and your prompts and insights are written from your own words by fixed rules; on, they're a little more on the money. Nothing else you write goes to any AI: not photos, not letters, not notes to a partner, not what the two of you share. Megan is still a written character: her lines are ours, picked by simple rules, and her voice is AI text-to-speech recorded in advance. If you turn on “Help build matching”, your pseudonymised answers (under a random id, never your name or email) may be used to train the models behind future matching.
If something you write names a day ahead, like “interview Thursday”, Megan asks how it went on that day. That's a simple rule that spots the word “tomorrow” or a day's name, on your phone's own words, and nothing else. Your answer stays in your journal.
The full statement is on the privacy and data use page.