Real-time transcription · Communication awareness

Speak.
Hear yourself.

ReEcho transcribes you as you talk, gently flags patterns that can quietly press on the other person — lecturing, over-reminding, commanding — and shows the exact words that triggered each nudge in your post-session report.

Get ReEcho → Read the privacy note

iOS Android Desktop coming soon

Live transcription 14:32 · Session active

How many times do I have to tell you —

just check the weather before you leave the house.

You never listen to a word I say.

I'm only doing this for your own good.

Just bring an umbrella next time…

This session · 2 nudges detected

Why ReEcho

It's not that the
other person
isn't listening.
It's that you
aren't hearing
yourself.

We tell ourselves we're saying it kindly. But what the other person hears may be a lecture, a verdict, or a quiet "I'm doing this for your own good."

Those signals feel natural coming out of your mouth. They sound foreign when you hear them played back.

ReEcho doesn't judge whether you're right or wrong. It simply hands your words back to you — so you can hear your own patterns, instead of the version of yourself you thought you were presenting.

Noticing is the first step to changing.

How it works

Four steps to
hearing yourself.

From starting a session to reading the report — clear, restrained, fully in your control.

01 / START

Good afternoon

SESSIONS

+ Start a session

RECENT

Yesterday · Family3 nudges

Monday · Team mtg5 nudges

Last week · Partner1 nudge

Start a session

Tap once. You're in.

02 / CONSENT

Explicit consent

Plain-language disclosure. Until you say yes, the mic stays off.

03 / LIVE

REC · 00:14:32

How many times do I have to tell you — just check the weather before you leave. I'm only doing this for your own good.

Live transcription

Words appear as you say them. Triggered phrases are gently highlighted in place.

04 / REPORT

Nudges this session

Preachy60%

Over-remind30%

Commanding10%

"I'm only doing this for your own good." — at 00:14:12

A report you can re-read

Counts, breakdown, the exact phrases that triggered, and what to try next time.

What it does

Four things that turn
a vague feeling
into something
you can read back.

Real-time transcription

While a session is running, every word you say is converted to text on the fly. Mandarin and English supported, with end-to-end latency under one second.

Communication pattern detection

Spots the phrases that often add pressure to a conversation — lecturing, over-reminding, commanding, moralising — and quietly marks them as you talk.

Reports you can re-read

After every session, a written summary: nudge counts, the patterns that came up, the exact words that triggered them, and one or two things to try next time.

Privacy you control

No mic until you agree. Stop a session at any time. Revoke consent in settings. Audio is never sold, never used for advertising.

Where it helps

Four kinds of
conversation —
all worth
hearing yourself in.

01 / INTIMATE

With your partner

Catch the "I'm only saying this for your own good" reflex that turns care into pressure. Put the concern back in the conversation.

02 / FAMILY

With your kids

Notice the constant reminders, the repeated lectures, the "because I said so." Save the teaching for later; keep the conversation for now.

03 / TEAM

With your team

Re-watch how you ran a meeting. Notice when your "feedback" sounded more like a verdict. Make an hour of talking worth an hour of listening.

04 / SOLO

With yourself

Turn "I think I come across fine" into "here's what I actually said." Practice out loud. Every sentence you speak deserves to be heard.

Privacy & data

Your voice.
Your call.

ReEcho uses cloud-based speech recognition for real-time transcription. Before a session starts, the app asks for your explicit consent. Only after you agree — and only for the duration of that session — is the microphone audio sent to Alibaba Cloud Intelligent Speech Interaction, where it is used to produce the live transcript. The text Alibaba Cloud returns is what ReEcho uses to detect patterns and generate the session report.

ReEcho does not sell audio data, and never uses it for advertising. You can stop any session at any time, and revoke consent at any point in Settings.

"If you don't agree, the microphone stays off." — that's a promise to every user.

Data flow · what happens during one session

You consent

Plain-language disclosure. You tap "agree."

Mic captures

Audio is captured locally, only while the session runs.

Alibaba transcribes

Encrypted audio is sent to Alibaba Cloud; it returns text.

Report is built

Only the transcript is used — for detection and your report.

FAQ

Things people
usually want
to know first.

No. ReEcho doesn't do "record and play back." It transcribes your speech in real time and uses that text to detect communication patterns. If what you want is to listen to your own voice later, that's not what ReEcho is for.

Only after you explicitly agree to start a session is microphone audio sent to Alibaba Cloud Intelligent Speech Interaction for real-time transcription. Alibaba returns the text. ReEcho does not sell audio data, and never uses it for advertising. When you stop a session, no further audio is sent.

Mandarin and English at the moment. The detection logic is tuned primarily for Mandarin-speaking users; English is supported and improving continuously.

During a session, tap Stop whenever you want — no further audio leaves the device. In Settings → Privacy you can withdraw your consent and request deletion of past session reports. You can also email support@reecho.chat and we'll take care of it manually.

No. ReEcho is a communication-awareness tool, not a medical, psychological, or therapeutic product. It doesn't judge you, and it doesn't promise to improve your relationships. What it does is simple: while you speak, it makes your words concrete and visible — so you can see for yourself which patterns show up most. Noticing is the first step to changing. The changing part is yours.

Get started