Workvivo experience

Workvivo is an employee-experience platform, and for organisations that have made it their internal home, it is where employees already go for news, recognition and company updates. Putting the assistant there means it sits alongside everything else employees check.

The trade-off is rendering. Workvivo bot chat displays plain text only — no markdown, no HTML, no native tables. Rather than leak formatting characters into the message, the assistant strips formatting and sends anything visual as an attachment.

For instance ID, API token and webhook configuration, see Workvivo setup.

How employees start a conversation

Employees message the Leena AI chat bot inside Workvivo. They are matched to their Workvivo user identity, so there is no separate login.

Only chat messages directed at the bot are handled. Other Workvivo activity — posts, comments, reactions — is not routed to the assistant.

How responses are formatted

Everything is converted to plain text before sending:

• Bold, italics, strikethrough, inline code, headings and blockquotes are stripped. Headings become ordinary lines.
• Horizontal rules become a simple separator line.
• Lists become plain bullets or numbers.
• Tables are rendered as styled PNG images, optimised for mobile reading, and delivered as attachments.
• Inline images in an answer are extracted and sent as attachments rather than left as raw markdown.

Responses are dispatched as ordered chunks — text first, then images — so a table arrives immediately after the paragraph that introduces it.

If a table or image can't be rendered or uploaded, the assistant falls back rather than failing: a table becomes a text list, and an image becomes its URL in plain text.

Interactive elements

Quick replies are sent as tappable reply options, and buttons render inside a card with either a label and a link, or a label the employee taps to send back as text. Because replies are matched on their text rather than on a hidden identifier, keep option labels short and distinct — two options with similar wording are harder to resolve reliably.

The 640-character limit is the tightest of any chat channel here, so long answers split across several messages more often than on Slack, Teams or WhatsApp.

Webviews are sent as links. Workvivo opens them in its own in-app viewer, and the assistant suppresses its own close button so employees use Workvivo's native one instead of seeing two.

Files and media

Outbound works well. The assistant downloads the file and uploads the bytes directly to Workvivo, rather than handing over a URL for Workvivo to fetch. Files are delivered through short-lived private links, so a URL copied out of Workvivo stops working shortly afterwards.

Inbound is not supported. Files an employee attaches to a Workvivo message are not processed, so the assistant responds as though nothing was attached. Voice input is not available either.

Notifications

Notifications arrive as Workvivo chat messages, so they inherit whatever Workvivo notification settings the employee has, including mobile push. There is no opt-in step and no messaging window.

Delivery and read receipts are not available on this channel. Employees who have never interacted with the assistant cannot be reached proactively.

Workvivo users are synced into the assistant so employees are recognised on first contact; that sync is part of channel setup.

Compared with the Web App

Things to tell your employees

• Answers here are plainer than in the app. The content is the same.
• Tables arrive as images, so they can't be copied or searched.
• Long answers split across several messages. The options appear on the last one.
• Attaching files won't work. Describe the issue in words instead.
• Tap a link to open it inside Workvivo; use Workvivo's own close button to come back.

Setting up the channel

Instance ID, API token, base URL, chat bot user ID, the webhook URL and user sync are covered in Workvivo setup.