Speaker identification
Timeglass puts names beside the conversation, so you can see who made a decision, asked a question, or committed to the next step.
From a voice to a name
Speaker diarization separates a recording into who spoke when. Speaker identification adds names to those turns. Timeglass brings that context into your transcript, timeline, and meeting notes, so you can check who approved the work and who agreed to follow it through.
Timeglass uses your microphone and the call audio to help distinguish your voice from people on the call. Meeting context helps it put names beside the words. You can read the transcript during the meeting and return to the finished record to correct a speaker’s name.
A teammate who enables voice recognition in Speaker Names can be recognized in later conversations. That available voice information helps Timeglass keep names consistent as you meet again. The setting gives each person control over whether their voice information is retained for recognition.
When several teammates record the same call, Timeglass can combine matching recordings into one conversation record. It can use one copy of overlapping audio as the basis for the transcript. Names give contributions context when you review a client request, an agreement, or a commitment after the meeting.
Select a speaker’s name in the meeting record to make a correction. The updated name helps readers follow the conversation and understand the action items that came from it. Short, unidentified speech can appear together under Other speakers, keeping the speaker view readable.
The shape of the conversation
Names and percentages sit above a colored timeline, with a lane for each speaker. Follow a lane to find that person’s contribution, then return to the surrounding conversation for context.
A speaker’s percentage is their share of attributed speaking time. Silence sits outside that total, while the timeline places the speech within the length of the meeting.
Return to the conversation
Select a timestamp or citation beside a decision or action item to return to that moment. The transcript shows the speaker, and chapters help you move between topics. Available audio and video let you review the conversation and what was shared on screen.
The work around the call
Speaker names help you follow decisions and commitments. Action items can include an owner, a due date, and a supporting moment when the conversation provides those details. Timeglass connects meeting records with project work and time spent preparing and following up. Review client and project assignments before the time reaches your report.
Through MCP, your AI assistant can use the meeting records and work history your account can access. Ask it to prepare the next agenda using the last call’s decisions and the work completed since, then review the draft.
Your questions
Timeglass separates speaker turns and adds names using the meeting context and available voice information. Your microphone and the call audio help it distinguish your voice from other participants. You can correct a speaker’s name in the finished meeting view when a label needs attention.
When a teammate enables voice recognition in their Speaker Names setting, Timeglass can use that voice information in later meetings. Recognition depends on the available audio and context. Your own microphone can still help identify your speech when that setting is off.
When teammates record the same call, Timeglass can combine matching recordings into one meeting record and use a single copy of overlapping audio. This helps keep the conversation together when several people on your team capture the call from their own computers.
Each percentage shows a speaker’s share of the attributed speaking time. The total covers speech assigned to speakers, with silence outside that total. The colored timeline shows when those contributions occurred during the meeting.
Select a timestamp or citation to return to its place in the conversation. You can follow the speaker-labelled transcript alongside available audio and video, or use chapters to move between topics. Playback depends on the recordings available for that meeting.
Connect Timeglass through MCP to let a supported AI assistant use the meeting records your account can access. Speaker names, decisions, and project history give the assistant context for preparing an agenda or drafting follow-up work. Review the result before sharing it.
Your next conversation
Try speaker identification alongside your meeting notes, project history, and reviewed time.
Try 14 days free