Meetings, transcribed live
Accurate transcription while people are still talking. German first-class, twelve languages covered.
Record, transcribe, summarise — nothing leaves unless you say so.
Every meeting becomes a decision document: recorded, transcribed, summarised — with action items, owners and deadlines. It runs on the device in front of you, so your recordings stay in the company. No server in between, no vendor holding your voice.
What you get / six surfaces
Most recorders give you text. Aurus gives you a decision document: who said what, what was agreed, who owns it — searchable across every meeting you ever ran.
Accurate transcription while people are still talking. German first-class, twelve languages covered.
Who said what — separated per speaker, and honest about the limits of the room.
Tasks, commitments and deadlines pulled out automatically — a list your team can execute, not a wall of text to re-read.
The same outcome rewritten for the board, the customer or the team — pick the tone, keep the facts.
Find the meeting where it was promised. Every session searchable — without anything leaving the house.
Original audio and structured summary stored side by side. What was said is provable.
Privacy — built in, not promised
Voice is the most sensitive data a company produces. Four commitments, built into the architecture — not into the terms of service.
Your keys stay yours.
Access credentials live in the device’s secure vault — never on our servers, never phoned home.
Works fully offline.
The built-in local mode processes speech on your own hardware. Flip one setting and nothing ever touches a network.
No middleman.
If you do use a cloud model, it streams directly from your machine. We are not in the path — there is nothing at nyxCore to subpoena.
Recordings stay in the house.
Sessions live on your device. Shared only when you decide, with access you control.
The best accuracy currently comes from the cloud path; one switch makes it fully local. We tell you that instead of hiding it.
Ships for Mac today; Windows and Linux follow. If your fleet is mixed, plan the rollout with us.
Speaker separation is only as good as the room. One microphone per speaker — we won’t pretend software fixes acoustics.
Early access is open. Pilot it in one team meeting this week — pricing and rollout planning go through the main page.