Voice input
Voice input records a short clip. Intra AI transcribes it and inserts the text into the message box. Review and edit the transcript before sending the message.
The microphone control appears only when voice input is enabled for your organisation and the browser supports audio recording. It may be unavailable in trial access.
The microphone control is in the message box; no recording is active
Example instance with sample data. Available controls depend on your role and configuration.
Dictate a message
- Open a chat and select Voice input beside the message box.
- Allow microphone access if the browser asks.
- Wait for Recording and speak clearly.
- Watch the level display and timer. If the display remains still, check the selected microphone.
- Select:
- Use recording to stop and transcribe it;
- Discard recording to cancel without transcription.
- While Transcribing… is displayed, wait for the result.
- Review the inserted text, correct names, figures and specialist terms, then send it.
If the message box already contained text, the transcript is appended. Check the join and punctuation. You can add several recordings in sequence.
Improve transcription
- Use a nearby microphone and reduce background noise.
- Speak at a natural pace with short pauses between thoughts.
- Say punctuation only when the intended structure would otherwise be unclear.
- Spell unusual names or correct them afterwards with the keyboard.
- Split a long dictation into shorter sections.
- Always check figures, email addresses, proper names and confidential details.
The language is normally detected from the recording, but your organisation may configure a fixed language. Quality depends on language, pronunciation, microphone and surroundings.
Manage microphone permission
The permission prompt belongs to your browser or operating system. If you denied access accidentally:
- Open site permissions for your Intra AI address.
- Set Microphone to allow.
- Check which input device is selected.
- Reload the page and begin a new recording.
Menu names differ between browsers.
What happens after an error
Discarding a recording inserts no transcript. If transcription fails, text already in the message box remains. Try recording again and review the result before sending.
Your organisation can set limits for recording size, duration and frequency. If a limit message appears, split the recording into shorter sections or wait before trying again.
Troubleshooting
| Message or problem | What to do |
|---|---|
| The microphone control is missing | Ask your administrator about feature access and use a browser that supports audio recording. |
| Microphone access denied | Allow the microphone in site settings and reload. |
| No microphone found | Connect or enable an input device and check system settings. |
| No input from microphone | Choose the correct device, check its input level and start again. |
| No recording captured | Speak only after recording starts, then try again. |
| No speech detected | Reduce noise, speak more clearly or use another microphone. |
| Transcription failed | Try later; contact your administrator if it keeps failing. |
| Processing does not finish | Wait briefly, reload if needed and check that existing text was retained. |
Privacy when dictating
A recording may capture more than the intended sentence, including background speech. Start only when ready and discard recordings containing unintended material. Consult your organisation’s privacy information for the processing and retention that applies to your instance; do not assume one universal setup.