Skip to main content
Audio Management controls how Text-to-Speech (TTS) output is generated, cached, and replayed. Without understanding this layer, teams often think changes “didn’t apply” when they are actually hearing cached audio. What caching is (and isn’t)
  • Cached audio stores previously generated TTS so it can be replayed instantly.
  • This reduces latency and keeps repeated phrases consistent.
  • Audio is only cached if the same utterance is generated at least twice within a 24-hour window.
  • One-off utterances will not persist in cache by default.
Do
  • Open Audio Management.
  • Review the list of cached utterances:
    • Greeting
    • Transfer / handoff language
    • SMS offer phrasing
    • Closings and confirmations
  • For any high-frequency utterance:
    • Open it and review how often it has been used.
    • Adjust stability and clarity for that utterance only if needed.
    • Use the play button to preview changes.
  • If an utterance must remain stable:
    • Generate it multiple times within 24 hours, or
    • Upload a static audio file to overwrite the cached version.
Interaction style (response latency) Interaction style controls how quickly the agent responds after detecting user speech. This directly affects interruption rate and perceived naturalness. Common modes:
  • Turbo (~400ms): Extremely fast, higher interruption risk.
  • Swift (~1200ms): Prioritises speed.
  • Balanced (~1600ms): Default for most use cases.
  • Precise (~2000ms): Slower, more deliberate, fewer interruptions.
Barge-in Barge-in determines whether callers can interrupt the agent mid-speech.
  • Useful for fast modes (Turbo/Swift).
  • Can feel chaotic if enabled without careful phrasing and latency tuning.
Verify
  • After any voice or phrasing change, start a new call session.
  • Confirm you are hearing updated audio, not a cached version.
  • Validate that turn-taking still feels natural after changing latency or barge-in.
[ ] Pronunciations Goal: Ensure domain-specific terms are spoken clearly and correctly in Call. Pronunciations are defined in Rules and applied globally. They modify how text is converted to speech, without changing the underlying text. When to use pronunciations
  • Brand names or product names that are mispronounced.
  • Proper nouns (locations, people, departments).
  • Numbers or IDs that need structured read-back.
  • Any phrase where pacing matters for comprehension.
How pronunciations work
  • Matching is done using regular expressions.
  • Replacements can be:
    • IPA (International Phonetic Alphabet)
    • SSML, such as <break> for pauses
    • Regex capture groups (\1, \2, etc.) for reformatting.
Examples IPA correction:
  • Regex: \bLouvre\b
  • Replacement: /ˈluːvrə/
  • Case sensitive: FALSE
Phone number formatting with pauses:
  • Regex: (\d{3})[ -]?(\d{3})[ -]?(\d{4})
  • Replacement: \1 <break time="0.5s" /> \2 <break time="0.5s" /> \3
Do
  • Add pronunciations incrementally.
  • Test each change in Call before adding more.
  • Prefer clarity over cleverness—overly complex regex is hard to maintain.
Verify
  • Mispronounced terms are corrected consistently.
  • Pauses improve comprehension rather than slowing the call excessively.