New 'Agentic Studio' Component for AI-Powered IVR Generation in wolkvox Studio

In wolkvox Manager, the new "Agentic Studio" component has been incorporated within wolkvox Studio. This component is designed for generating IVR flows using AI based on natural language instructions. This functionality transforms the way voice routing points are designed, allowing users to describe what they need in functional terms and letting the system translate those requirements into a usable technical structure within the flow.

The component is located in the "Cognitive" components group and is currently only available for voice (IVR) routing points. It is not enabled for chat routing points or other types of flows. From its configuration panel, users have a conversational AI experience where they can type prompts directly or even dictate them using the microphone button for voice-to-text conversion.

The functionality is based on progressive interaction: the designer starts with an instruction about what they want to build, and from there, the AI acts as a technical copilot that interprets the requirements, identifies information gaps, and asks additional questions to complete the flow logic. This way, it not only generates a structural proposal for the IVR but also builds a more robust technical result in parallel.

As a result, Agentic Studio can generate optimized PHP code, facilitating the creation of menus, wait times, DTMF captures, validations, confirmations, transfers, and other typical IVR behaviors. This also helps in scenarios of technological evolution, such as migrating traditional tone-based flows to more advanced experiences involving components like ASR or TTS.

In the "Chat" tab, the conversation history with the AI is displayed, along with the field to write each prompt and the microphone button. As the conversation progresses, the AI responds with questions or proposals to complete the required logic. Subsequently, the user can review the generated result in the code view and finally confirm the process with the "Save and Close" button.

Benefits

• Allows creating IVRs from natural language instructions.
• Significantly reduces the technical development time for voice flows.
• Acts as a technical copilot to translate functional requirements into implementation logic.
• Generates optimized and readable PHP code.
• Helps reduce syntax errors and logical failures in flow construction.
• Facilitates the creation of menus, validations, confirmations, and transfers without starting from scratch.
• Adds value in migrating traditional IVRs to more advanced AI-powered experiences.
• Allows using voice-to-text to dictate prompts via the microphone button.
• Improves time-to-market in designing and adjusting IVR solutions.

Configuration

To use Agentic Studio in wolkvox Studio, follow these steps:

1. Open the component's configuration panel by double-clicking on it.
2. In the "Chat" tab, write the initial instruction describing the IVR you want to build.
   • Press "ENTER" to send each prompt when finished.
   • If you prefer, use the microphone icon button to dictate the prompt via voice.

Continue the conversation with the AI by answering the questions the system asks to complete the flow logic.

Once the AI reviews and understands everything and is ready to generate the code, the "Code" tab will open directly with all the logic built by the AI.

1. Review the generated result in the "Code" tab.
   • Carefully validate the technical proposal and the generated code.
2. When you are finished, click "Save and Close."

Considerations

• This component is only available for voice (IVR) routing points.
• It is not enabled for chat routing points.
• Each prompt is sent by pressing the "ENTER" key.
• The use of AI, ASR, or TTS components within the generated flow may involve additional consumption of credits or tokens, depending on the current pricing.
• It is essential to review and validate the generated code and logic before finalizing the design.

Redesign and New Configuration Options for Web Chat Widgets in wolkvox Manager

In wolkvox Manager, the configuration section for web chat widgets has been completely redesigned to offer a clearer, more modern, and flexible management experience. This update reorganizes the widget editing interface and expands the possibilities for visual and functional customization, allowing better adaptation of the web chat to the needs of each implementation.

Now, when editing a web chat configuration within Omni Channel > Web Chat, the user finds a renewed structure with tabs such as "Configuration," "Design," "Conversation," and "Form," along with a side preview that allows reviewing how the changes affect the widget.

Within the "Configuration" tab, new fields have been incorporated. One of them is "Phone Field Limits," located in "Basic Configuration," where minimum and maximum character values can be defined for the phone field in the initial chat form. This validation helps better control the data entered by the customer before being transferred to a human advisor or a chatbot.

Also in the "Configuration" tab, the new section "Behavior on Load" has been added, allowing users to choose between two widget deployment methods:

• Floating button, so the user can click to open the chat.
• Automatically expanded chat, so the widget appears open when the site loads.

Additionally, the "Design" tab has been added, which incorporates new options for customizing the overall appearance of the widget. This includes configurations such as "Main Colors" and "Typography," with options to modify header color, header text color, chat background color, and typography. The "Buttons" section has also been added, where background, text, and border colors can be defined for both normal and hover states.

The new "Conversation" tab expands the customization of the messaging experience. It includes:

• "Message Layout," with options for "Conversation Style" (recommended) or "Left."
• "Bubble Styles," to define background and text colors for both customer and advisor/bot bubbles.
• "Identification in Bubbles," to configure how the customer's, advisor's, and bot's names are displayed.
• "Profile Image," to set images via public URL for customers and advisors/bots.
• "Message Area," with options such as voice notes, attachments, send icon style, and chat language.

This update also includes the "Request Data Before Chat" option, which allows users to decide whether the initial form will be mandatory or not. If configured as non-mandatory, the customer can start the chat immediately upon opening the widget. This option will only be available when only the Chat checkbox is active in Configuration > Channels. If other channels such as WhatsApp, audio call, video call, or callback are also enabled, this configuration will not appear.

Benefits

• Offers a more modern, organized, and user-friendly configuration interface.
• Improves the usability of the web chat widget editing process.
• Expands the visual customization options for the widget.
• Allows better adaptation of the chat experience to each company's graphic identity.
• Facilitates controlling the validation of the phone field from the initial form.
• Provides flexibility to decide whether the chat opens with a floating button or automatically.
• Improves the visual presentation of messages with new conversation and bubble styles.
• Allows clearer identification of who sends each message.
• Makes it possible to customize profile images when the chat layout allows it.
• Allows adjusting message area functions such as voice notes, attachments, send icon, and language.
• Facilitates configurations where the customer can start the chat without filling out a form when only the chat channel is used.

Sending WhatsApp Sticker Messages from the "Respond Chat" Component in wolkvox Studio

In wolkvox Manager, it is now possible to send WhatsApp stickers automatically from the "Respond Chat" component in wolkvox Studio, enriching the conversational experience within the flows configured in Routing Points.

With this new feature, the "Respond Chat" component incorporates the ability to attach a sticker resource via a public URL, as long as the file is in the appropriate .webp format. This allows the flow to deliver more expressive visual content directly in the WhatsApp conversation, in addition to the configured text message.

As seen in the configuration, the sticker is defined in the attached file link field within the component. Once the flow is saved and executed, the sticker is automatically sent to the end user in the WhatsApp chat. It can also be viewed in the conversation from wolkvox Agent and in the history review from Data Monitor, providing traceability of the visual content sent.

This enhancement expands the expressive capabilities of bots and automations, enabling the creation of more dynamic, engaging, and brand-aligned interactions.

Benefits

• Allows enriching WhatsApp flows with automated visual content.
• Makes it possible to send stickers from the "Respond Chat" component without additional development.
• Provides a more human and engaging communication by conveying emotions or visual tones.
• Facilitates the use of custom stickers to reinforce branding, mascots, phrases, or brand elements.
• Enhances the conversational experience by making messages more attractive and memorable.
• Can help increase user engagement and retention within the flow.
• Maintains traceability of the sent sticker in wolkvox Agent and in chat reviews from Data Monitor.

Configuration

To use this functionality, follow these steps:

1. Go to "wolkvox Studio" and open the chat-type "Routing Point" where you want to configure the sending.
2. Open the configuration panel of the "Respond Chat" component by double-clicking on it.
   • In the component configuration, locate the "Attached File Link" field.
     • Paste the public URL of the sticker there.
     • Verify that the file is in .webp format.

In the Data Monitor conversation history, you can also view the sent stickers.

Considerations

• The sticker must be hosted on a publicly accessible URL.
• The required format for the file is .webp.
• The functionality applies to sending from the "Respond Chat" component within wolkvox Studio Routing Points.
• The sent content can be viewed later in both wolkvox Agent and Data Monitor.

New Fast, Intelligent, and Reasoning AI Selector in the MCP Component of wolkvox Studio

In wolkvox Manager, a new configuration capability has been incorporated for the MCP component in wolkvox Studio: the AI model selector with Fast, Intelligent, Reasoning, and Automatic options.

This functionality allows flow designers to assign the most suitable type of AI engine based on the cognitive load required for each interaction within the MCP component. From the component configuration, it is now possible to define whether a task should be resolved with a speed-focused approach, greater analytical capacity, or deeper reasoning, depending on the nature of the prompt or base instruction being processed.

The logic behind this improvement allows working with a kind of intelligent model routing, where the system can operate with different AI profiles to better balance response speed, analytical precision, and cost efficiency in credit consumption. This makes it possible to reserve more robust capabilities for complex tasks while using lighter models for simple or low cognitive demand requests.

Additionally, this evolution also allows customizing technical parameters independently, such as temperature and token limit, giving greater control over the behavior of the model used in the MCP component. Thus, the designer can adapt each flow not only to a functional need but also to operational efficiency and cost optimization criteria.

Benefits

• Allows selecting the most suitable AI profile based on task complexity.
• Helps optimize costs by using lighter models for simple queries.
• Reserves greater reasoning capabilities for complex, technical, or logical cases.
• Improves the balance between speed, precision, and credit consumption.
• Reduces latency in simple interactions.
• Provides more control over the behavior of the MCP component.
• Allows for a more strategic configuration of AI-powered flows within wolkvox Studio.
• Facilitates a better end-user experience by responding more quickly when deep analysis is not required.

Configuration

To use this functionality in the MCP component of wolkvox Studio, follow these steps:

1. Open the component's configuration panel by double-clicking on it, and locate the AI Model section.
2. Select one of the available options:
   • Fast.
   • Intelligent.
   • Reasoning.
   • Automatic.

Usage Recommendation

• Use "Fast" for simple queries or low-complexity tasks.
• Use "Intelligent" for intermediate tasks that require better interpretation.
• Use "Reasoning" for complex, technical, or deep logic cases.
• Use "Automatic" when you want the system to determine the best alternative based on the complexity of the prompt.

Consideration

Credit consumption will depend on the model used in each case, so this configuration should also be evaluated from a cost and performance perspective of the flow.

New Embedded Audio Playback in Speech Analytics within "Voice of the Customer"

In wolkvox Manager, a new embedded audio playback capability has been incorporated within the "Voice of the Customer" section, allowing users to listen to recordings directly from the browser without needing to leave the platform or open external tools.

This improvement is available in the "Dashboard" > "Voice and Text Analytics" > "Voice of the Customer" path, where the detail of each interaction now includes a "Play Audio" button that activates an integrated player in the same view. When clicked, the system connects with the "Core" recordings to enable real-time streaming of the audio associated with the interaction.

In addition to playing the audio, the interface displays a sound wave synchronized with the recording, allowing users to accompany the listening with the analysis already available on the screen, such as the summary, justification, overall sentiment of the conversation, and interaction-related metrics. This way, the user can review both the audio evidence and the text and sentiment analysis generated by the platform in a single panel.

This update strengthens the use of Speech Analytics by more directly integrating call listening with the analytical context of the interaction, facilitating a more agile and centralized review.

Benefits

• Allows listening to recordings directly from wolkvox Manager.
• Avoids leaving the browser or relying on external players.
• Reduces auditing and interaction review time.
• Centralizes audio, text analysis, and Speech Analytics metrics in a single panel.
• Facilitates validating the real context of the conversation while reviewing sentiment and justification.
• Enhances the analysis experience by displaying an audio wave synchronized with playback.
• Makes consulting audio evidence within "Voice of the Customer" more efficient.

Visualization

To use this functionality, follow these steps:

1. In wolkvox Manager, go to the left sidebar menu and navigate to "Dashboard."
2. Open the "Voice and Text Analytics" tab.
3. Go to the "Voice of the Customer" section.
4. Select the interaction you want to review. In the detail panel, click the "Play Audio" button.
5. The system will display the embedded player at the bottom and begin streaming the associated audio.
   • Use the available controls to play, pause, stop, or adjust the listening speed as needed.