This guide walks you from a fresh laptop to a completed recording — install, connect, record, review. If you already have a Treasure AI Voice account and just need the installer, jump to Download.
- A Treasure AI Voice user account (if you don't have one, ask your administrator — see Account Activation).
- A supported computer (see system requirements).
- Microphone access permission on macOS or Windows.
Download the appropriate build from the Download section on the overview page.
- Open the downloaded
.dmgfile. - Drag the Treasure AI Voice icon into the Applications folder.
- Eject the DMG and launch Treasure AI Voice from the Applications folder or Launchpad.
- On first launch, macOS may prompt you to allow the app to run — click Open.
- Run the downloaded
Treasure-AI-Voice-Setup-*.exe. - Follow the installer prompts. The default location and shortcuts are recommended.
- Launch Treasure AI Voice from the Start menu.
On Windows, the app checks for new releases in the background and installs them the next time you quit and relaunch. On macOS, install updates by downloading and re-running the DMG.
On first launch, the app shows a two-step setup wizard.
- On the Welcome step, click Continue.
- On the Connect step, click Connect Treasure AI Voice. Your default browser opens the Treasure AI Voice sign-in page.
- Sign in with the same method you use for the mobile app or the web admin dashboard (Sign in with Google, or Email / Password).
- When the browser confirms the connection, return to the desktop app. The wizard advances to Done and the Meetings page opens.
For details on sign-in methods and supported identity providers, see the Administrator Guide.
If the Connect button returns an error, verify that outbound HTTPS to *.treasure.ai is not blocked by your VPN or corporate firewall. Contact your administrator if the block is intentional.
The desktop app has two surfaces:
| Surface | Purpose |
|---|---|
| Status Popover (menu bar / system tray) | A small window that opens when you click the app icon in the macOS menu bar or Windows system tray. Use it to start, pause, resume, and stop a recording without switching windows. |
| Meeting Manager | The main window. Shows your recordings list, transcripts, summaries, action items, and settings. Open it by double-clicking the tray icon, or by choosing Open Meetings from the tray's right-click menu. |
Right-clicking the tray icon opens a menu with Open Meetings, pause / resume / stop actions while a recording is in progress, the current app version, Check for updates… (Windows only — macOS updates by re-running the DMG), and Quit. The version entry is handy for confirming which build you're running before reporting a beta issue.
Closing the Meeting Manager window hides it to the tray — it does not quit the app. To fully quit, right-click the tray icon and choose Quit.
The first time you start a recording, the operating system asks for microphone access.
- macOS: Allow access when prompted. If you missed the prompt, open System Settings → Privacy & Security → Microphone and enable Treasure AI Voice.
- Windows: Allow access when prompted. If you missed the prompt, open Settings → Privacy & security → Microphone and enable Treasure AI Voice under Let apps access your microphone.
Click the tray icon to open the Status Popover, then click Start Recording.
The Recording Consent dialog opens.
Read the consent script aloud to every participant and wait for their verbal consent. Then:
- Choose a summary template for this recording (you can change the default in Settings; see Summary Templates).
- Click I have consent — start recording.
The Meeting Manager shows an Active Recording banner with a live mic-level indicator, a timer, and Pause / Resume / Stop controls.
The mic-level bar turns green as your voice is detected. If it stays flat while you're speaking, the wrong input device is selected — see Troubleshooting.
When the meeting ends, click Stop. The recording moves to the queue and uploads in the background.
Every participant must give verbal consent before you start recording. The consent script in the dialog is the exact wording legal has approved. This applies to all Treasure AI Voice recordings — mobile, desktop, and PLAUD device — and is logged locally as part of the recording's audit trail.
Once the upload completes, the recording appears in the Meetings list. Click it to expand the detail panel.
You'll see:
- Summary — an AI-generated summary in the template you chose.
- Action items — extracted tasks and next steps.
- Transcript — the full transcript with speaker labels and timestamps.
Each section has a Copy button that copies the raw markdown to your clipboard.
Transcription usually finishes within a few minutes for meetings under an hour. Longer recordings take proportionally longer. If a recording gets stuck in the Failed state, click Retry on the row to re-upload from the local copy (available until the local retention window expires — see Settings below).
Open the Meeting Manager and click the Settings icon in the left sidebar.
Notable settings:
| Setting | Description |
|---|---|
| Local recording retention | How long the app keeps the original audio file on your computer after upload. Default is 1 day. Increase this if you expect to retry failed uploads; decrease it to free disk space sooner. Cloud retention is set separately by your administrator (7 years by default for compliance). |
| Transcription language | Auto (recommended), English, or Japanese. Auto lets the transcription service detect the language of each recording. |
| App language | English or Japanese. Follows the OS locale on first launch. |
| Appearance | Light, Dark, or System. |
| Legal | Links to Privacy Statement, Terms of Service, Voice Supplemental Terms, AI Terms, and AI Acceptable Use Policy. |
The wrong input device is selected. The desktop app auto-selects a working input device on launch, but plugging in a headset or switching monitors after launch can leave a stale selection.
Fix: Stop the current recording, quit and relaunch the app, then start a new recording. The app re-picks the best available input on launch.
The upload didn't complete — usually a transient network issue. Click Retry on the recording row. The app re-uploads from the local audio file (available until the retention window expires).
If retry keeps failing:
- Open Settings and click Check connection under the account section.
- Confirm outbound HTTPS to
*.treasure.aiis allowed on your network. - Retry is only offered for rows in the
FAILEDstate. If a recording is stuck in another state (for exampleUPLOADING), leave it — the app will finish the upload or move it toFAILEDautomatically.
Tokens are stored encrypted per user on the machine. If you sign into a different macOS/Windows user account, or wipe the app's data directory, you'll need to reconnect. Data directory locations:
- macOS:
~/Library/Application Support/Treasure AI Voice/ - Windows:
%APPDATA%/Treasure AI Voice/
- Summary Templates — customize what your summaries look like.
- Administrator Guide — how organizations manage users, devices, and recordings.
- Troubleshooting & FAQ — full troubleshooting reference (mobile + desktop + hardware).
Feedback on the beta? Contact your Treasure AI account team.