Product Update - v5.6, Webex Connect Bot Builder Platform, October 2023

Webex Connect Bot Builder v5.6 release brings multiple new capabilities to help you continue delivering seamless customer experiences. Here’s a list of the key updates:

• Addition and modification of new/old templates in Apple messages for business
• Addition of a new multilingual model for training - Polymatch
• Addition of Email and Phone number entities
• Limited cross-launch of the platform from the control hub
• Option to select Vector model in Swiftmatch engine
• Deprecation of WeChat as a channel
• Validation for voice settings in response designer
• Support for additional Webex Engage VA regions
• No integrations in Webex Connect for new bots
• Update operation disabled for Task bot responses with errors

Please refer to the details below to learn about all the changes and enhancements.

📘

Release Date

The date of release will be communicated separately over an email.

Details

Changed - Quick Replies in Apple Messages for Business (AMB)

Apple recently made some changes to its quick replies payload. While the functionality remains the same, the UI has been updated to match the nomenclature used by AMB. Here are the changes that we made –

• 'Text’ label was updated to ‘Summary text’
• Payload’ label was updated to ‘Identifier’

Changed - Media template in Apple Messages for Business (AMB)

According to Apple’s update for supported media types in Apple messages for business, the template type in the bot platform has been updated. Here is the list of supported media types –

• Plain text
• HTML
• JPEG
• PNG
• MPEG
• OGG
• MP4
• PDF

Added - Forms template in Apple Messages for Business (AMB)

Business Forms Messages empower users to design intricate, multi-page interactive experiences for iOS and iPadOS devices using a unified JSON payload. Given the powerful and extensive customizations that Apple allows for forms, we chose to go with JSON to better cater to the client’s business needs. This feature enables businesses to gather extensive customer data through a user-friendly interface integrated within the Apple Messaging Interface. This seamless interaction allows for diverse engagement options without necessitating users to exit the conversation. Sending the structured content templates (body and metadata) illustrated in the sample JSON will allow a bot platform user to send Business Forms to the consumer device.

Each JSON will contain information about the pages that will go as a part of the form with customizations. The pages can be of different types –

• splash: An introduction page that includes a title, description, action button, and an image.
• select: A page that features either a single “True/False” selection or multiple select. The page also supports image assets with each selection.
• picker: A page that features a single select from multiple options.
• datePicker: Provides Apple’s standard Date Picker with the ability to change the date format and set maximum/minimum dates.
• input: A page that can feature various input fields to collect strings of information.
  The structured content metadata empowers the user to specify the form's pages and visual layout. This Metadata template comprises two sections -
  • BusinessFormEvent – Information like page identifier for the starting page, summary, pages, images can be configured.
  • BusinessChatMessage – received and reply message information can be configured.

Added - Polymatch for training multi-lingual bots

To enhance the bot builder platform’s multi-lingual capabilities in Q&A bots, we have introduced our new multi-lingual model ‘Polymatch’ as an option for training bots. This is available as a part of the Swiftmatch NLU engine’s advanced engine settings. For a bot trained using Polymatch, a distinct model is created for each language added to the bot and inference is performed based on the language of incoming user query. The existing language-agnostic model is available as a Language encoder in the dropdown. Since Polymatch is not language agnostic and trains the corpus for each language, it requires significant training data for each language, unlike Language Encoder. Please note that the languages supported by Polymatch and the existing language encoder are different. Please refer to the platform documentation for more information.

Added - Email and Phone number entities

Predefined system entities allow developers to detect and extract common types of data from end-user inputs. Based on the platform usage, intents often require user information like email and phone number. To preclude developers from creating regex entities to extract this information, email, and phone numbers will be available as system entities.

Email: entity type can be used to gather email addresses following the common @ convention.
Phone number: entity type can be used gather phone and mobile numbers. These numbers can be with or without country codes. This will suffice for most use-cases where phone number is required. However, if some use-cases require phone numbers in a format that is not detected by the system phone number entity, developers can create a regex entity with the appropriate pattern.

Added - Cross launch from control hub

Bot builder is currently only accessible as an app in Webex connect app tray. With this release, developers interested in creating bots for voice channel will be able to launch Bot Builder from Webex Control Hub.

Clicking on the Webex connect Bot Builder link will redirect the user to platform dashboard. Users will be able to see Q&A bots and task bots on the dashboard. In addition to these, users will also be able to see Knowledge section in the navigation menu on the left. Knowledge section can be used by users interested in creating Knowledge bases for Agent Assistance (limited availability) use-cases. Like bot nodes in Webex connect, developers can select the bot they want to use in their voice flows in the Webex Contact Center Flow Designer. Only bots with voice channel configured will appear in the Virtual Agent dropdown in the flow designer. The default bot configured in the Bot Builder enterprise profile will appear in this dropdown for new flows. Developers can also launch Bot Builder from Flow Designer by clicking on ‘Configure selected Virtual Agent’.

Added - Vector model selection for Swiftmatch

Users can now select their preferred vector models as a part of the advanced engine settings in the Swiftmatch NLU engine. Selection is possible between two options – Utterance level vs Article level vectors. In our continued push to improve the accuracy of our NLU engines, we experimented with using article-level vectors as opposed to the older model of using utterance-level vectors and found article-level vectors improve accuracy in most cases. Please note the article level vectors will be the new default value for vectorization for new single-lingual bots and for multilingual bots article level matches will be supported only when the multi-lingual model is Polymatch.

Changed - WeChat deprecated as a channel

The bot builder platform, following in steps with the Webex Connect platform has made the strategic decision to deprecate WeChat as a communication channel. This means that WeChat will no longer be supported as a messaging channel. As technology and user preferences evolve, the platform is continuously evaluating its supported channels to ensure the best user experience and compatibility. While this change might impact users currently relying on WeChat, the platform aims to focus its resources on enhancing other communication channels, providing users with more robust and efficient options for building and deploying their bots.
Bots that have WeChat channels responses configured will need those channels to be removed and bots to be retrained. Sessions with WeChat will be retained for the 30 days sessions expiration period.

Added - Validation for voice settings in response designer

Voice as a channel was introduced in v5.4. Along with the channel, voice settings were introduced to handle the intricacies that come with voice as a medium. In this release, validations are added for these settings to prevent users from adding invalid values. The validations added are:

• Timeout (s): must be an integer between 1 and 30.
• Timeout between two digits (s): must be an integer between 1 and 30.
• Termination character: must be a single character between 0 and 9, * or #.
• Max length: must be an integer between 1 and 30. These fields in voice settings will now be covered by error navigation in the response designer. Developers can navigate to fields with invalid values by cycling through errors highlighted above response designer.

Added - Support for more Webex Engage VA regions

Prior to this release, the Service Region field in Webex Engage VA settings had only two valid values – ‘uk’ and ‘usa’. This meant that the Virtual Agent framework was only supported for Ireland and Oregon regions of engage. With this release, each bot builder environment will be mapped to a corresponding Webex Engage region and users will no longer have to provide a service region. For example, bots built in the US deployment of bot builder will work as a Virtual Agent for US deployment of Webex Engage only. This same will be applicable for Ireland, London, Canada and Sydney deployments.

Changed - No integrations in Webex Connect for new bots

With the release of updates Q&A and Task bot nodes built using the pre-built integration framework, the list of bots in the node UI is populated based on the bots the developer has access to. Integrations in Connect are no longer used to populate this list. Due to this, no integrations will be created in Webex Connect for new bots that are created.

Changed - Update operation disabled for Task bot responses with errors

Any draft of responses in task bots that contains errors can no longer be updated. Saving responses with errors leads to errors down the line that can leave end-users puzzled. To prevent this, developers will only be able to update responses once they resolve all the errors in responses. Developers can easily cycle through the errors by clicking on the error navigation arrows above the response designer.

Changelog