Live Transcription
Swiftscore automatically transcribes your classroom observations so you can stay focused on what's happening in the room instead of taking notes by hand. This article covers how transcription works during a recording, how to label speakers, how to upload audio for transcription, and what happens to your transcript afterward.
How it works during an observation
When you start recording an observation, Swiftscore begins transcribing the audio in real time. As the lesson unfolds, you'll see:
- A live audio waveform so you can confirm your microphone is picking up sound.
- The transcript building line by line as speech is captured.
- A word count so you can track how much has been captured during the session.
You don't need to do anything to start transcription beyond starting your recording — it runs automatically in the background.
Capturing context before you start recording
If you type observation notes before pressing record, that text is captured as context for the session. Your typed notes appear above the live transcript rather than being overwritten once recording starts, so any context you jot down beforehand (lesson objective, focus area, etc.) stays attached to the observation.
Speaker labels
Swiftscore distinguishes between different speakers in the recording (for example, teacher versus students) and displays a label next to each line of the transcript.
If a label isn't quite right, you can click it to rename it — for example, correcting a label to read "Teacher" or a specific name. Your renamed labels are remembered for that observation.
Filtering the transcript
Once segments of the transcript have been classified, you can filter the view — for example, to look specifically at teacher questions versus student responses. This makes it easier to review a recording for a specific pattern (like how often the teacher asks open-ended questions) without scrolling through the entire conversation.
Classified segments may take a short time to appear after you finish recording — if you don't see filtering options immediately, give the transcript a moment to finish processing.
Uploading audio for transcription
In addition to live recording, you can upload an audio file (for example, a recording made on another device) and Swiftscore will transcribe it for you. Drop the audio file in during the observation creation flow, and Swiftscore will process it and merge the resulting transcript into your observation.
Uploading other files
You can also attach supporting files to an observation — lesson plans, student work samples, photos, and similar documents. Swiftscore checks for duplicate uploads so you don't accidentally attach the same file twice.
Notes
Alongside the transcript, you can type your own notes during or after an observation using the built-in notes editor. The editor supports basic formatting — bold, italics, and bulleted or numbered lists — using standard keyboard shortcuts (Cmd/Ctrl+B for bold, Cmd/Ctrl+I for italics) or the toolbar buttons.
Saving and reliability
Your transcript and notes are saved automatically as the observation progresses — you don't need to manually save. Swiftscore continuously syncs the transcript in the background, including safeguards so that if you navigate away or close the tab unexpectedly, your captured transcript is preserved rather than lost.
After the observation is complete, you can return to it to review the full transcript alongside your notes and any uploaded files.