Skip to main content

January 25, 2025

This release brings new AI powered conversational bots, additional reports and others!

Change log​

  • NEW Voice conversational bots
    • We’re introducing voice conversational bots to address the limitations of DTMF-based workflows and enable more advanced scenarios, such as answering customer queries using knowledge base searches. The new node, called Conversational Bot, can be added to a voice workflow and includes a welcome message (text) that plays when customers call, along with an instructions field where designers specify the bot’s purpose and style of response. Users can also select a language for speech-to-text processing (English, Spanish, Serbian, Hungarian, or German) and define dynamic outputs for different outcomes, such as transferring a call to a live agent
  • NEW Email conversational bots
    • We’re introducing email conversational bots as a new node in the email workflow, which allow users to define free-text instructions outlining the bot’s purpose, capabilities, and response style. The node includes an email template selector that uses an existing template to construct email replies, with a new template variable that can dynamically insert the bot’s response. Once the bot sends the email, the workflow proceeds under an Email Sent outcome, and users can create dynamic outputs for scenarios like transferring to an agent if the customer requests human assistance.
  • NEW Chat conversational bots
    • We’re introducing a new Conversational Bot workflow node for chat, which offers an instructions field where you can specify its purpose, response style, and any limitations or rules it must follow. You can also select an existing knowledge base and grant access to various third-party APIs. Additionally, the node provides dynamic outputs, allowing you to define the names and conditions under which each output should be used, giving you greater flexibility and control over the bot’s behavior.
  • NEW Conversation categorization rules for voice
    • We’re introducing the ability to configure automatic conversation categorization rules for voice channels through the portal, allowing administrators to assign a single categorization rule to each voice channel. When the conversation ends and the full call transcript becomes available, the system will automatically apply the corresponding tags based on these rules, using LLM to process the transcript.
  • NEW Conversation categorization rules for voice
    • We’re introducing conversation rating insights in the portal under a new Insights menu group, accessible separately for voice, chat, and email. Each media-specific menu will display a rating report containing a pie chart that illustrates the distribution of ratings. Users can filter these insights by date range, providing a clear view of customer feedback trends over time.
  • NEW Filter conversation reports
    • Users can now filter historical conversation reports by media type, channel, workflow, date, or conversation length (above or below a specified threshold), making it easier to locate and analyze specific interactions.
  • RESOLVED Fixed an issue where portal might display empty rows when scrolling quickly through the historical reports.
  • UPDATED Agents can no longer cancel a break or break reservation by clicking on the same break name. Agents must now specifically click on the Available entry in the profile menu to cancel a break reservation. This is based on user feedback as it was discovered that agents tend to cancel their pending break reservation accidentally by double-clicking the break name. The leave break option from the task area is unaffected by this change.
  • RESOLVED Fixed an issue where forwarding emails would result in the attachments beind duplicated.
  • RESOLVED Fixed an issue where scrolling through an email thread containing multiple emails, with one email open, would make the scrolling automatically jump to the top of the thread.