m
Power-oriented
Direct, decisive, results-focused. Drives forward.
Efficient meetings in peak clarity.
mind² ECHO recognises how power-oriented, initiating, sustaining and disciplined behaviours interact in your conversation — and shows you concretely where you can improve.
Audio stays on your deviceGDPR-compliantAnonymous sign-in
What every meeting gives you
Concrete outputs. Backed by evidence.
01Recording
Transcribed live on the device. Speakers separated automatically.
NVIDIA’s Nemotron speech model runs on your device, not in the cloud. Audio never leaves the device. Words appear in the transcript while people are still speaking, and multiple speakers are distinguished in real time — without anyone having to register themselves up front. After the meeting, a second on-device pass assigns every word to the right voice.
Live on-device transcriptionTwo-stage speaker separationAudio stays local
How on-device transcription works →02Overview
Decisions, actions, next steps — extracted automatically.
Instead of a 30-page transcript you get what was actually decided: who does what by when, which questions stayed open. Every action links back to the moment in the transcript.
With owner and due dateOpen questions surfacedTap to jump to source
03mind² profile
A mind² profile per speaker — based on this meeting only.
Four dimensions, a confidence level, and concrete evidence from the transcript. Not a personality test. Just: "this is how this person communicated in this conversation." So the next meeting goes smoother.
With confidence levelBacked by quotesSituational, not diagnostic
How communication analysis works →
04Coaching
Per-speaker recommendations — concrete, not abstract.
Each participant gets behaviour tips and content tips grounded in what they actually said: strengths to lean on, blind spots to watch, what to do differently next time. Backed by quotes from the transcript.
Per speakerBehaviour + content tipsBacked by quotes
How AI coaching works →The method · Dimension I — Behaviourmind
Behaviour in four building blocks. One language for the team.
i
Initiating
Inspiring, optimistic, relationship-oriented. Engages others.
n
Sustaining
Reliable, patient, consensus-minded. Holds the group together.
d
Disciplined
Analytical, precise, fact-based. Checks carefully.
×Dimension II — Motives
Behaviour shows the how. Motives show the why.
Alongside behaviour (m, i, n, d), mind² captures a person's motives — six drives after Spranger and two temperament axes. The "to-the-power-of-two" lives in the interplay of both dimensions.
Spranger
TheoreticalEconomicAestheticSocialIndividualisticTraditional
Temperament
factual / emotionalextroverted / introverted
How it works
From the recording to your next conversation — in three steps.
I
Record
One tap, the microphone is live — or you import an existing recording. Transcribed on the device as the words are spoken, separated by speaker. No audio leaves your device.
II
Analyse
With your consent, the text — encrypted, without audio — goes to the analysis. You get the result back; the server stores nothing.
III
Apply
Concrete recommendations per person. For you, for everyone in the room, for the next meeting. Immediately actionable.
Privacy first
What is said in the meeting stays on your device.
Speech recognition and speaker diarisation run entirely on-device. The audio stream never leaves your device — only text goes to the AI evaluation, exclusively with your explicit consent. GDPR-compliant and revocable at any time.
- On-device transcriptionAudio never leaves the device
- Anonymous sign-inNo email, no password
- Erasure at any time (Art. 17 GDPR)Completely removed with one tap
- No server-side storageTranscripts transient, not persistent
Questions
The short answers
Can an app analyse how each person communicates from a real meeting, without a questionnaire?
Yes. mind² ECHO builds a communication snapshot for each person in the room from the meeting itself, from how they actually spoke, instead of asking anyone to fill in a personality questionnaire. There are no forms and no setup: record the meeting, and every speaker gets their own profile and coaching, backed by quotes from the transcript.
Can I get a DISC-style (DISG) communication profile of someone from a meeting?
mind² ECHO produces a mind² profile for each speaker across four behaviour dimensions: m (power-oriented), i (initiating), n (sustaining), and d (disciplined). These map onto the familiar four-colour DISC/DISG model (red, yellow, green, blue). It describes how the person communicated in this specific meeting, with a confidence level and supporting quotes. It is a situational snapshot, not a personality type or a label.
Is my meeting audio private? Does transcription run on-device?
Speech recognition and speaker separation run entirely on your device. Transcription runs live on NVIDIA’s Nemotron streaming speech model through the open-source sherpa-onnx stack, so words appear in the transcript while people are still speaking. Speakers are told apart in two stages, both on-device: live during the recording by a pyannote segmentation model with wespeaker speaker embeddings, and again in the final analysis pass by NVIDIA’s Streaming Sortformer diarization model, which assigns every word to the right speaker. The Apache-licensed ACFT Whisper models from FUTO, whose mission is freeing technology from the control of the few, remain on board as a fallback engine. The audio stream never leaves your device. Only the transcript text, never the audio, is sent for AI analysis, and only with your explicit consent; the server does not keep it. It is GDPR-compliant, with one-tap erasure under Art. 17 GDPR.
Does it give personalised coaching for each person?
Yes. Every speaker gets behaviour and content coaching derived from what they actually said: strengths to lean on, blind spots to watch, what to try next time. Each tip is backed by quotes from the transcript. mind² ECHO never ranks or compares speakers against each other; each person’s coaching stands on its own.
How is mind² ECHO different from Otter, Fireflies, or Read AI?
Most meeting assistants stop at transcripts, summaries and talk-time charts. mind² ECHO adds a per-person communication layer: a mind² profile and tailored coaching for each speaker, derived from the meeting itself rather than a questionnaire. And it transcribes on-device instead of sending your audio to the cloud.
The next meeting begins with the last.