Genesys Cloud - Main

 View Only

Sign Up

  • 1.  Introducing Custom Vocabulary Control for Genesys Extended (EVTS) Transcription (API-Based)

    Posted 03-19-2026 10:06

    We're excited to introduce a powerful new capability that gives you more control over how your transcripts look and perform-managing custom dictionary terms for the Genesys Extended (EVTS) transcription engine via API. 

    What is this feature? 

    This feature allows you to create and manage custom vocabulary terms that improve transcription accuracy for your organization's unique language-like product names, acronyms, or industry jargon. 

    With this, you can: 

    • Improve recognition of domain-specific words 

    • Control how terms appear in transcripts (e.g., "covid" → "COVID-19") 

    • Add acronyms and multi-word phrases 

    • Fully manage terms via API (create, read, update, delete) 

    • Track when vocabulary updates are applied 

    How it works 

    You interact with the Dictionary Feedback API: 

    POST /api/v2/speechandtextanalytics/dictionaryfeedback  

    Each term you add becomes part of a custom vocabulary used by the Genesys Extended transcription engine in your Organization, for the given language/dialect you target. 

    Key things to know 

    1. Required fields 

    When creating a term, you must include: 

    • term → The spoken word or phrase (max 30 chars) 

    • dialect → e.g., fr-FR 

    • transcriptionEngine → must be "GenesysExtended" 

    Optional: 

    • displayAs → how it appears in transcripts (max 50 chars) 

     

    2. displayAs = formatting power 

    This is where the feature shines: 

    Spoken Term 

    displayAs Output 

    covid 

    COVID-19 

    genesys 

    Genesys 

    four one k 

    401k 

    Coca Cola 

    Coca-Cola 

    If omitted, transcripts default to the original term value.   

    3. Important validation rules 

    • No digits in term (must spell out numbers) 

    • No hyphens in term (use spaces instead) 

    • No leading special characters 

    • Use displayAs for formatting like hyphens or numbers 

    4. Acronyms require special syntax 

     

    To define acronyms: 

    • Use periods between letters: F.B.I. 

    • Include a trailing period 

    • Example #1: 

    • "term": "F.B.I." 

    • "displayAs": "FBI" 

     

    • Example #2: 

    • "term": "Dynamo D.B." 

    • "displayAs": "DynamoDB" 

     

    • If your acronym is plural, you must use a hyphen between the acronym and the 's' 

    • Example #3: 

    • "term": "A.B.C.-s" 

    • "displayAs": "ABCs" 

    Note: Not all dialects support this syntax (e.g., Japanese, Chinese, Arabic).   

    5. Status lifecycle (important for expectations) 

    After creating or updating a term: 

    Pending → Active  

    • Changes typically take ~15 minutes to apply 

    • Updates or deletes return to Pending while rebuilding 

    • Failed status means the vocabulary update didn't apply 

     

    API Examples 

    Create a term 

    POST /api/v2/speechandtextanalytics/dictionaryfeedback 
     
    { 
      "term": "covid", 
      "dialect": "fr-FR", 
      "transcriptionEngine": "GenesysExtended", 
      "displayAs": "COVID-19" 
    }  

     

    Acronym example 

    { 
      "term": "A.T.M.", 
      "dialect": "fr-FR", 
      "transcriptionEngine": "GenesysExtended", 
      "displayAs": "ATM" 
    }  

     

    Update a term 

    PUT /api/v2/speechandtextanalytics/dictionaryfeedback/{id} 
     
    { 
      "term": "covid nineteen", 

    "dialect": "fr-FR", 

    "transcriptionEngine": "GenesysExtended", 
      "displayAs": "COVID-19" 
    }  

    List terms 

    GET /api/v2/speechandtextanalytics/dictionaryfeedback?transcriptionEngine=GenesysExtended  

     

    Delete a term 

    DELETE /api/v2/speechandtextanalytics/dictionaryfeedback/{id}  

     

    When should you use this? 

    This feature is especially useful if: 

    • Your transcripts include brand names or internal terminology 

    • You need consistent formatting across transcripts 

    • You want to improve accuracy for acronyms or jargon 

    • You operate in industries with specialized vocabulary 

     

    Final thoughts 

    Custom vocabulary is a simple but powerful way to: 

    • Boost transcription accuracy 

    • Improve readability 

    • Tailor transcripts to your business language 

     

    Start small-add a few high-impact terms-and iterate as you go. 

    If you have questions or want help getting started, don't hesitate to post them in this thread.  

    Regards,


    #WEM-Quality,WFM,Gamification,etc

    ------------------------------
    Jose Ruiz
    Genesys - Employees
    Product Manager
    jose.ruiz@genesys.com
    ------------------------------


  • 2.  RE: Introducing Custom Vocabulary Control for Genesys Extended (EVTS) Transcription (API-Based)

    Posted 04-07-2026 09:35

    Interesting feature but when it comes to administering/maintaining such a dictionary, i understand that it is not so convenient. There is not any UI we could use to manage that. Every action (append terms, modify terms, delete terms) is done via invoking the respective API calls. If that is the case, it is too bad. It would be way more efficient to have such dictionaries being displayed directly in Genesys Cloud UI. Are there any plans to support this as well?



    ------------------------------
    Orestis Dimitropoulos
    ------------------------------



  • 3.  RE: Introducing Custom Vocabulary Control for Genesys Extended (EVTS) Transcription (API-Based)
    Best Answer

    Posted 04-07-2026 22:18

    This is a really strong enhancement for Genesys Extended, especially for acronyms, brand names, and business-specific vocabulary.

    One point I'd add is that, in real customer scenarios, the newest STT engine is not always the one that performs best for every language or accent variation. In some cases, other engine paths may still deliver better recognition results.

    That is why this release is great news, but it also opens an interesting discussion about whether dictionary support could eventually be extended more broadly across other STT engine options as well.



    ------------------------------
    Mateus Nunes
    Tech Leader Of CX at Solve4ME
    Brazil
    ------------------------------



Related Content