# Speech-to-Text (STT)

> Speech-to-text (STT), also called automatic speech recognition, is technology that converts spoken audio into written text. It is used live, for example so a voice agent can understand a caller, and on recordings, for example to transcribe meetings and sales calls. Many STT systems also label different speakers, add punctuation and timestamps, and handle multiple languages.

## Why it matters for your business

Transcription is the first step in most automations built on calls and meetings, such as CRM notes, call coaching, compliance review and meeting summaries. Errors at this stage carry through everything downstream, so accuracy on your own audio matters more than published benchmarks. Names, product terms, accents and poor phone audio are the usual trouble spots, and many services let you supply a custom vocabulary. Recording and transcribing calls also brings consent and retention obligations, which vary by state and by industry.

## Example

A B2B sales team records discovery calls through its video conferencing tool. After each call, an automation sends the recording to a speech-to-text service with speaker labels, then passes the transcript to a language model that extracts budget, timeline, decision makers and next steps into the CRM opportunity. Reps review the fields instead of typing notes from memory.

Related terms: https://www.lumeor.studio/glossary/text-to-speech/, https://www.lumeor.studio/glossary/voice-ai-agent/, https://www.lumeor.studio/glossary/latency/, https://www.lumeor.studio/glossary/structured-outputs/

---

Source: https://www.lumeor.studio/glossary/speech-to-text/