Digiotouch AI for Linux

Notes de version et historique

Version 2.013 mai 2026

Highlights

First public release of Digiotouch AI for Linux. A native desktop app that records your meetings — both sides — and turns them into AI summaries, transcripts, and action items. Built specifically for Linux with Wayland-native screen capture, PipeWire audio, and a portable AppImage you can run without installing.
AI meeting recording and summarization. Record any meeting with system audio (the other side) plus your microphone, and get an AI-generated summary, transcript, action items, key topics, and conclusions automatically.
Multiple recording sources. Capture a specific application window, a browser tab, or your full screen, with optional microphone audio for narration. Powered by xdg-desktop-portal for Wayland-native screen sharing and a Rust GStreamer pipeline for reliable MP4 output.
Live captions and live translation. Real-time captions appear in the right-hand sidebar while you record. Switch the target translation language mid-call and the app reconnects automatically. Final transcripts and translations are saved as text files next to your recording.
Four-language interface. Use Digiotouch AI in English, French, Japanese, or Estonian. Switch from the language picker in the top bar — every screen, menu, button, error message, and notification is fully localized. Your meeting content (summaries, transcripts, action items) remains in the meeting's original language.
Dark mode. A full dark theme designed for low-light work and long meeting reviews. Choose light, dark, or follow-system from Settings → Appearance. The choice persists across launches and applies instantly to the entire app.
Native desktop notifications. Get a Linux notification (via libnotify) when a meeting summary is ready, when a recording finishes uploading, or when something goes wrong. Permission-aware: the app asks once and remembers your answer. Each notification type (summary ready, upload complete, recording failed) can be toggled independently from Settings → Notifications.
Auto-update. The app checks for new versions on launch and every 24 hours after that. When an update lands, a banner appears in the top-right with Install Now, What's new, and Later. Updates are cryptographically signed (Ed25519 / minisign) and verified before install. AppImage users get one-click self-update; .deb users update through apt.
Privacy & data controls. Two explicit toggles in Settings → Privacy & Security. Crash & error reporting (default ON) sends anonymous error reports to help us find and fix bugs — no recording content, no transcripts, no personal data. Product analytics (default OFF) sends anonymized usage data so we can prioritize what to build next — same privacy guarantees. EU data region for both.
Recording library. Browse all your past meetings on a single dashboard. Search by name, sort by date, and filter by status. The list grows with your window — zoom out or maximize and you'll see more meetings without empty space at the bottom.
Folders and tags. Organize recordings into folders by project or team, and tag individual meetings for cross-folder grouping.
Editable summaries. Refine any section of the AI-generated output — title, summary, action items, conclusions — directly from the recording detail page.
Sharing. Copy a shareable link to any recording, export a PDF report, or send the summary to Google Docs.
Expanded integrations catalogue. Communication: Slack, Microsoft Teams, Discord. Email: Gmail, Outlook Email. Knowledge & docs: Notion, Confluence, Google Docs. Project management: Jira, Asana, Linear, Monday, Trello, ClickUp. CRM: Pipedrive, Salesforce, Hubspot, Attio, Freshsales, Zendesk, Apollo.io, Zoho. Storage: Box, Dropbox, SharePoint. Telephony: Aircall. Webhook / no-code: Clay, Zapier. More connectors are actively in development.
Sign in with Google, Apple, or Microsoft. Authenticate with your existing account, or use email + password if your organisation prefers. Switching accounts on the same machine clears all previous-account data cleanly — no bleed-through of recordings, summaries, or preferences.
In-app diagnostic export. Need to send our support team your app version, OS, and environment in a single paste? Settings → About has a one-click "Copy diagnostic info" button that bundles everything support needs into your clipboard.

Linux specifics

Two distribution formats. Pick whichever fits your setup. AppImage is portable, single-file, and needs no install — download, chmod +x, double-click to run; auto-updates in-place. .deb is a system install for Debian/Ubuntu derivatives, bundling GStreamer plugins, the Yaru icon theme, and a shared-MIME-info update so the app integrates cleanly with your desktop; updates flow through apt.
Wayland and X11 supported. Native screen capture uses xdg-desktop-portal so the OS handles the "share this window" picker — same flow as Firefox, OBS, and other well-behaved Linux apps.
PipeWire and PulseAudio. Audio capture works on both stacks. System audio comes from the PulseAudio monitor source on your default sink (PipeWire's pipewire-pulse compat works identically).
Custom window chrome. The title bar is rendered by the app rather than the window manager, so it matches the app's appearance across GNOME, KDE, and tiling WMs without per-WM theming.

System requirements

64-bit Linux (x86_64). Tested on Ubuntu 22.04+ and 24.04+, Debian 12+, Fedora 40+.
Wayland or X11 session.
WebKit2GTK 4.1 (libwebkit2gtk-4.1-0) and GTK 3 (libgtk-3-0) — pulled in as dependencies by the .deb; bundled or assumed present for the AppImage.
PipeWire or PulseAudio for audio capture.
xdg-desktop-portal plus a backend (xdg-desktop-portal-gtk or xdg-desktop-portal-kde) for screen capture.
Network access for sign-in, cloud summarization, and integrations.

Languages

🇬🇧 English (en)
🇫🇷 French (fr)
🇯🇵 Japanese (jp)
🇪🇪 Estonian (ee)
Switch any time from the language picker in the top toolbar.