/

Speech & Audio

/

Automatic Speech Recognition (ASR)

Automatic Speech Recognition (ASR)

/ automatic-speech-recognition-asr /

Technology that converts spoken audio into text. The transcription layer of every voice agent, typically measured by word error rate.

Technology that converts spoken audio into text. The transcription layer of every voice agent, typically measured by word error rate.

Why it matters

Every downstream decision starts from the transcript. If ASR mishears an account number or intent, no amount of clever prompting recovers — so ASR accuracy bounds agent accuracy.

Related — Speech & Audio