Week of April 13
Support for Viber connector, third-party bots, secure forms and social messaging
Features
Agent Avatar
Consumers will see the agent's avatar during a conversation with a brand on Viber. In case no avatar is set, Viber falls back to the image of the Viber channel.
Handle unsubscribe events
Consumers can unsubscribe from a brand's channel on Viber. In case this happens during a conversation, an agent will no longer be able to send messages to the consumer. With the new release, the connector detects this event, sends a notification to the agent and closes the conversation.
Features
New connector type - Custom Endpoint
Enables brands to plug in an orchestration layer between their bot and the Conversational Cloud. Moreover, it enables AI vendors to provide a suitable API to integrate into LivePerson's ecosystem.
Allow brands to provide a custom URL as an endpoint for our connector. If a bot is configured in Third-Party Bots pointing to this URL, our connector will call this URL with the consumer message as the payload. The service behind this URL is expected to return the bot response in a predefined format.
Enhancements
Custom Integration vendor rebranding
The Functions as a service AI Vendor in Third-Party Bots is renamed from “Custom Integration” to “LivePerson Functions”. This was done to separate the Liveperson Functions connector from the “Custom API Endpoint” connector, which was introduced with this release.
Preselect Agent as the bot type on new bots
Previously, while creating a new bot using the Third-Party Bots wizard, the user had to select the bot type manually. Now, on the first step of choosing the bot type, “Agent” will be selected by default.
Survey bots - optional skill field
Users no longer have to select a skill from the list of skills when survey bots are configured in Third Party Bots. Now, with no skill selected, the bot will subscribe to all incoming surveys associated with the selected App ID.
Enhancements
- Increasing Character Limit
- Secure forms had the limitation that consumers can submit only 1900 characters and this character includes both questions & responses. We have now increased the character limit up to 7000.
Features
Support character limit for Tweets
Agents can send public messages (tweets) within the character limit of 280 that is standard for the platform. Longer tweets can still be typed out and a visual guide shows overflow of characters, allowing for informed editing. The message cannot be sent if it is over the character limit, as seen in the UI via the disabled send button.
Invite to Direct link not counted in Tweet's character count
Agents can add an Invite to Direct link in their public replies without the link taking up precious character space. In the UI, the button is the same as before, except now it 'turns on' to indicate the addition of the link to the reply, rather than getting included in the text itself. The invite is automatically created as a button on Twitter's native platform and this detail can also be seen in the transcript area of the New Agent Workspace under the agent's message metadata.
Enhancements
Preview of Twitter videos in New Agent Workspace
Twitter videos and gifs shared by consumers will now be directly embedded in the transcript area of the New Agent Workspace, improving preview and support pertaining to such content.
Support actions (View Detail and Reply) on public images
The enhanced Agent Workspace now allows performing actions (view details and reply) on public images directly from the transcript.
Fixes
Remove X button from input area
As seen in all environments, when the agent clicks on ‘reply’ on messages sourced from Twitter and then cancels it using the X button in the input area, the Public tag disappears and only reappears by switching to Direct or another conversation in the list.
We have found the use and behavior of the X button redundant and to resolve this bug and create a better experience for the agent, this button has been removed from the UI.
Show metadata on messages generated from different accounts on Twitter
Not all messages in the transcript area have the metadata shown under each agent to consumer message. This is particularly problematic when the agent is switching the account to reply from, and back to back messages sent from different accounts do not show the correct metadata under them.
Since this fix, metadata wouldn’t be shown under a message only in these scenarios:
- Multiple messages are sent back to back from the same account, in the same channel (i.e. Public or Direct)
- Bot messages