# How to work with transcript records
For users of TrueConf Server, working with the TrueConf AI Server solution involves recognizing conference audio tracks and using the obtained transcripts (scripts). In this case, the video communication server administrator can configure automatic audio recognition at the end of the event, so manual initiation will not be required.
The following instruction is intended for TrueConf Server users who will create conferences, manage their recognition, or simply access transcripts.
You can work with the list of transcripts in your personal account on TrueConf Server and in the personal account on TrueConf AI Server. The interface itself is similar and provides the same capabilities. To learn how to work with conferences in the video communication server's personal account, read its documentation.
Below is how to work with the transcripts list in the TrueConf AI Server web interface.
# Activation of recognition in conference settings
For TrueConf AI Server to be able to convert audio to text, you need to activate the necessary parameters in the conference settings (it looks the same in the server's personal account and in the TrueConf client application scheduler):
When creating a conference, go to the General tab.
Check the box Create a conference transcript (it may be enabled by default if your administrator has activated automatic recognition for all conferences).
Select the primary language that the AI will attempt to recognize audio in first. If there are phrases in another language, they will also be correctly recognized, but the default language will assist the AI in cases where similar languages are used.
# Transcript list
The administrator of TrueConf Server could configure chat notifications about the conference being ready for recognition and the transcript being ready. In this case, you will receive a link to the transcript in such a message.
Otherwise, you can find the conference transcript directly in its card in your personal account or the client application's scheduler:
You can also simply go to your TrueConf AI Server address and complete the authorization process:
Go to the main page of your MCU server (check the link with your administrator).
On the page that opens, select your video communication server from the dropdown list and click Sign in:
Enter the login (TrueConf ID) and password for your video communication server account and click Sign in:
If you are logging in for the first time, you will need to confirm the permissions for your account to the AI server, just click Allow:
After that, a list of transcripts available to you and conferences ready for recognition will open. For more details on access, read further:
Each conference entry in the list is marked with one of the statuses: recognized, in the process of recognition, or not yet recognized. By clicking on the row of a recognized conference, you will be taken to the transcript page.
To start transcribing an unrecognized conference, click the 
button.
To download the text of an existing transcript, click the 
button. You will be able to choose the export format for the transcript: as a csv table, a txt text file, or a pdf document. The transcript can also be downloaded as an archive with the original audio recording of the conference.
You can delete the transcript by clicking the 
button. This will remove both the transcript and the original corresponding conference audio track.
To quickly find the necessary transcripts, you can use filtering by conference name or ID, duration, event time, and the status of the transcript itself.
In the shared list, you can perform a comprehensive search through transcripts for specific text entries. Initially, a list of matching transcripts will be displayed, and upon opening a specific transcript, the search will continue within it with the relevant lines highlighted.
# Transcript access settings
A user can be assigned one of two access levels to the transcript, and this will determine the list of features:
If you have full access, you will be able to: create notes (if such a module is activated on the AI server), download, configure access for other users, and delete.
if you have read access, you will only be able to download the transcript.
Who initially has access to the transcript depends on the settings on the TrueConf Server where the event took place. The administrator might have configured the access rights as follows:
Disable access for all users (including owners) - this will prevent you from viewing transcripts from this server, even for your own events;
enable access for moderators only;
grant full access to the moderator, and read-only access to other participants;
grant all participants full access.
Additionally, the administrator of TrueConf Server could configure access for specific users. For example, a group of users with full access to all transcripts and their summarization can be created. These users will then have access regardless of the specific transcript settings. If you are administering the video conferencing server, read more about the settings in its documentation.
Access to a specific transcript is granted by a user who has full access to it.
If you have full access to the transcript, you can configure access for other users, including those from another video communication server if your AI server is integrated with it. To do this, click the 
button and configure access:
The access level for each user is selected from the drop-down list (read or full). To revoke a participant's access to the conference, click the 
button. To add a user from your own or an external server, find them or enter the full TrueConf ID and click the 
button.
# Transcript page
On the page of a specific entry, you can:
review the text of the recognized conference;
start summarizing or view it if it has already been received;
listen to the original audio in the built-in player.
Clicking on any line will take you to the desired spot in the player, allowing you to immediately listen to the original.
To view the transcript as a plain text with timings (similar to the structure of a subtitle file in movies), go to the Script tab.
# Uploading Files for Transcription
In addition to native integration with TrueConf Server / TrueConf Enterprise, users of your AI server will have the ability to upload any file with an audio track for transcription. For instance, you can upload a video recording of a past conference where transcription was not enabled but was recorded. The number of speakers will be automatically identified based on voice differences, and the statements will be attributed to these participants of the event.
Supported for upload:
audio files: wav, mp3, flac, ogg, aac
Video files: mp4, avi, mov, mkv, webm.
Maximum file upload size: 5 GB.
The process of automatically identifying multiple different speakers within a single audio stream and segmenting the speech according to these participants is called diarization.
To upload a file, click the Upload file button in the transcripts list:
A download window will open, where you can:
Upload a file by clicking on the area listing formats, or simply drag and drop the desired file from the file explorer.
Specify the event name, otherwise the file name will be used.
The language for speech-to-text recognition is usually determined automatically with fairly high accuracy, but you can manually specify the preferred primary language.
If you know the number of speakers in advance, you can specify them to help distribute the cues more accurately.
After specifying all the parameters, click the Start button:
Once the file is fully uploaded to the AI server, speech recognition and speaker allocation will begin immediately.
Once the file is recognized, the transcript will appear in the general list. Click on it to edit the names of the speakers, which are labeled by default as Speaker 1, Speaker 2, and so on.
If you are familiar with the participants, you can recall who is who by their initial remarks and assign them the appropriate names. To rename the speakers, click the button above the transcript. This will open a window for configuring access and displayed names for the participants.
Click on the 
button next to the desired speaker and enter the required name. Then press the Enter key or click anywhere on the screen with the mouse — the changes will be applied immediately. Adjusting the names of speakers does not affect access permissions; you will need to configure everyone in the list above who should have the ability to work with the text.
If, for any reason, a single person appears as two speakers in the transcript (for example, due to a change in microphone that made the voice sound different), you can assign the same name multiple times in the list to consolidate the remarks under one participant. In other words, unique labels are not a mandatory requirement.
It is important to remember that renaming speakers is only available for transcripts from files and only after accessing the transcript itself.
← System