Skip to content
EN
Open app

For AI assistants: this page as plain markdown is at https://docs.voicelogica.ai/raw/agent-configuration/transcription.md, and https://docs.voicelogica.ai/llms.txt lists the docs by topic, one file each, small enough to read whole. Read those rather than this HTML page.

How the agent hears

4 min read
Copy as Markdown
View as MarkdownOpen the raw .md file.md

There is no Transcription item on the editor rail. Hearing settings live on Identity - Voice & Language under How the agent hears callers, plus Call handling - Advanced for talk-over and keypad.

The banner on that block: tweak these only if calls feel laggy or the agent interrupts too often. Defaults work for most use cases. Manage reusable vocabulary on the Keywords pages (linked from that banner).

Voice and Language - How the agent hears callers
accuracyExpand

The post-call transcript on Calls is a result. It is not configured on a hidden Analysis Agent. When a recording has two or more distinct caller voices, each caller line carries a Speaker 1, Speaker 2 (and so on) badge so you can tell the speakers apart.

First - name the symptom, then open one place

Section titled “First - name the symptom, then open one place”
What you heard on Try this agentWhere to fix itWhat happens if you skip it
Missed a brand name, street, or SKUVoice & Language - Custom vocabulary, and/or company Keywords / keyword setsThe agent keeps hearing nearby words and answers the wrong thing.
Talks over the caller, or stops on a TVAdvanced - interruptions and Voice Sensitivity. AdvancedSliders here will not help if the real issue is Skip Turn.
Cut off while they typed digitsAdvanced - keypad wait / end keyThey never finish the account number.
No audio to replay, but the summary existsRecording & Privacy - recording was off. RecordingNot a hearing problem.
Summary or properties missingCall results, not hearing. Call resultsNot a hearing problem.
  1. Reproduce the miss

    Try this agent and say the brand / street / SKU the same way a caller would.

  2. Start on Voice & Language

    Under Accuracy, use Custom vocabulary for uncommon words, names, or terms. Type a word and press Enter or comma.

    Below it, Transcription context (Soniox) takes a plain sentence about what callers will say, for example that they state a 9-digit tax number. Numbers are then transcribed as digits instead of times or dates.

    Send what the agent says (Soniox) passes the agent’s own words to speech-to-text while it speaks. Short replies to its questions (yes/no, one of the options it offered, the digits it asked for) are then heard in context, and the agent’s voice leaking back on the line is not taken for the caller. It uses the newer Soniox model and applies from the next call.

  3. For reusable sets, open Keywords

    Follow the Keywords link on the hearing banner. That opens the company Keywords area (not a rail item named Transcription).

    Create a New keyword set with:

    Details
    Name and language for the set.
    Phrases
    Add phrases one by one (+ Add) or Bulk add. Each phrase can have a Boost value.
    API sync
    Optional sync controls when you manage phrases from outside the UI.

    Agents reference these sets from their hearing / transcription settings. Cancel if you were only exploring - empty sets are fine until you have real brand terms.

    New keyword set Phrases tab
    accuracy-keywordsExpand
  4. Try this agent again with the same word

    If it is still wrong, the set is on a different agent, or the caller’s wording does not match what you stored.

Some plans show extra turn-taking or transcription-provider controls under Voice & Language. If you do not see them, your plan uses the defaults - that is normal. Do not hunt for a Transcription tab on the agent rail.

If the agent talks over digits, that is Skip Turn, not a hearing slider.

Where is the Transcription tab in the agent editor?

There is no Transcription tab on the editor rail. Hearing settings are under Identity - Voice & Language, in the How the agent hears callers section, and reusable vocabulary lives under Keywords.

I added a keyword, so why does the agent still mishear our street name?

Either the keyword set is on a different agent, the caller says it differently from what you stored (for example the old street name), or you only typed it in Custom vocabulary on one agent. Add that spoken wording to the keyword set too, then test again with Try this agent.

Your First Agent (Voice & Language) - Advanced - Recording & Privacy