You shouldn’t have to start from the beginning every time.
Turn on long-term memory and Melyvi holds your story between conversations — the people in it, the patterns it has noticed, the things that actually helped. Every line is traceable to your own words, and yours to correct, hide, export, or erase.
Free to start · Long-term memory is off by default · Adults 18+
The same sentence, to something that knows you and something that doesn’t.
“She hasn’t replied for six hours and I’m getting really anxious.”
Explains anxiety. Accurately, probably. To a stranger.
“You described something very like this after two earlier situations, and both times the not-knowing seemed to land harder than a clear answer — is that what’s happening again?”
It can only say that because two earlier moments are on file with their own evidence. A pattern needs at least two; one moment is not a pattern at all.
Three steps, and then it’s just talking.
Talk the way you'd talk to a friend
Type or speak. Choose one of five ways to be met, and change it in the middle of a conversation if it isn't landing.
It keeps only what it can point back to
People, episodes, emotions, patterns, what helped. An episode keeps the trigger, the thought, the reaction and the outcome — not a summary.
Next time, it already knows where you were
No recap, no re-explaining the same difficult person. Talk, remember, connect, notice, reflect — every conversation makes the next one better.
One conversation. This is what stays.
Melyvi reads back what you said and keeps the few things worth keeping. Every line has a source, and a guess never pretends to be a fact.
“The call with my mother left me feeling small again. I didn’t want to explain myself from scratch, so I went quiet and walked around the block instead.”
Watch the memory fill in →
Having to re-explain yourself often seems to come before you go quiet.
You say how you want to be met. It follows.
Five styles, switchable in the middle of a conversation. These are the actual instructions each one carries — being advised at when you wanted to be heard is a real thing, so the difference is written down rather than hoped for.
“Reflect back what you heard and stay with it. Do not analyse, advise, or reframe. At most one soft invitation to say more.”
“Use emotional labelling and reflective listening to help them name what is happening and why it might be happening.”
“Use Socratic questioning. Follow the thread underneath what they said rather than the surface content.”
“Prioritise emotional regulation. Slow the pace, shorten sentences, orient to the present. No analysis until they are steadier.”
“Use CBT-style questioning to test assumptions and predictions, warmly and without winning the argument.”
Every reply also carries the reason behind it, in one plain sentence — “why I’m asking this” is a button, not a mystery.
Quiet on purpose.
Structured self-help with clear boundaries.
Melyvi is the language layer around an approved self-help protocol. It can reflect and personalise the wording; it cannot invent a therapeutic procedure. Scores monitor change — they do not diagnose. You choose every exercise, action, and share.
Orient to the present
2–4 minutesReduce overwhelm enough to choose what to do next.
Slow the pace
1–3 minutesOffer a short regulation option when arousal blocks thinking.
Facts and predictions
5–10 minutesSeparate observed evidence from an anxious prediction.
Make room for a thought
3–7 minutesReduce struggle with a thought that cannot be solved right now.
One useful action
Plan in 3 minutesReconnect with routine, mastery, pleasure, care, or another person.
Make the problem smaller
5–12 minutesMove from rumination to one controllable next step.
Each use records the exact version delivered. A tool that could increase distress says so before you start, and says to stop if it does.
PHQ-9 and GAD-7, every 14 days
Spaced to reduce burden. They track whether anything is moving. They are not a diagnosis, and no number decides anything on your behalf.
One small behaviour, not more time in the app
Plan the smallest useful action, name the likely obstacle, keep a smaller backup. Afterwards you review mastery and connection — not whether you did it right.
Built before a crisis, not during one
Yours to write and yours to change. Crisis resources are served live, so a number can be corrected without waiting for an app release.
A clinician summary you read first
Build a preview, read exactly what it contains, then authorise it — or keep it as a preview and send nothing. An authorisation can be revoked afterwards.
Every message and new journal entry is screened first. An urgent turn skips the conversational model entirely and routes to local crisis and emergency resources.
The log keeps the route and the detector version. Never your words.
Privacy here is behaviour, not copy.
Long-term memory is a separate, optional choice, and it is off until you turn it on. Three switches each gate a real capability — not a display preference.
Having to re-explain yourself often seems to come before you go quiet.
The learned item and its stored evidence links are deleted. The original conversation stays unless you delete that separately.
Nobody is automatically reading your conversations.
Melyvi is an AI tool for reflection and structured self-help. It does not diagnose, replace professional care, or provide emergency monitoring. The audit trail records that a safety screen fired or a memory was forgotten — never what was said.
How care and safety work →Before you download.
No, and it is not allowed to be marketed as one. Melyvi is an adult general-wellness and structured self-help product — it does not diagnose, treat, cure, or clinically determine whether you are safe. Plenty of people use it between sessions, and it can help you arrive with the words already found.
The first conversation is the only one you’ll have to explain.
For adults 18+ · Long-term memory is optional and off by default · Store links are placeholders until the apps are listed.