Real-time transcription
Transcribing audio as it is spoken, with very low latency, rather than processing a complete recording after the fact (batch transcription).
Real-time (or streaming) transcription converts speech to text live, word by word, as someone speaks — the technology behind live captions and meeting assistants that show notes during a call. It trades a little accuracy for very low latency.
Batch transcription, by contrast, processes a finished recording and can spend more compute for higher accuracy and cleaner diarization. Building a private knowledge base usually uses batch transcription, since the recording already exists.