Ravenna AI in Slack

Configure AI features in Slack including automated answers, request type classification, and auto-tagging

Ravenna AI integrates with Slack to provide automated answers, intelligent routing, and other AI-powered features. You can customize which features your teams use and configure them to match your workflow.

Answers

Overview

Ravenna AI can respond to and provide answers to your employees’ questions directly in Slack. To get started, navigate to Knowledge and ensure that you’ve ingested some documents into your knowledge base. This information is what Ravenna AI will use to generate answers, so the more relevant content you have, the better the answers will be.

Once you’ve imported your existing content, you can enable answers in any of your connected channels. Navigate to your desired Queue settings page and toggle on Enable Ravenna AI.

You’ll receive a confirmation message in Slack letting you know that Ravenna AI is ready to help in the queue.

Multimodal Support

Ravenna AI now supports processing images alongside text messages in Slack. When users share screenshots, diagrams, or other visual content, Ravenna can analyze the images and provide contextual responses.

Image processing works with common formats (JPEG, PNG, GIF) and integrates seamlessly with text-based conversations. The AI can reference both the image content and accompanying text to provide comprehensive answers.

This multimodal capability is particularly useful for:

• Technical support scenarios with error screenshots
• Troubleshooting visual issues or interface problems
• Analyzing diagrams, charts, or documentation images
• Processing visual content from mobile devices or desktop captures

Images are automatically processed when included in messages, requiring no additional configuration from users or administrators.

Slack Answers

Once Ravenna AI is enabled, depending on your configuration, Ravenna will either respond to all messages and threaded messages in the channel, or it will respond only to messages posted in the channel.

You can also toggle on the setting so that Ravenna only responds to messages that mention @Ravenna. This is a great option if you want to give your team more control over when they want to use Ravenna AI.

When a user sends a message that Ravenna AI can answer, you’ll see robot emoji reaction on the message the user sent. This lets your employees know that Ravenna is working in the background and processing their request. Once Ravenna has generated an answer, it will post the response in a thread under the user’s message and the 🤖 reacji is removed from the message.

Ravenna will always send answers into the thread. This keeps channels clean and uncluttered.

Answer Actions

All answers from Ravenna come with a source document link and user feedback buttons.

The source link shows users exactly where Ravenna found the information, so anyone can dive deeper if needed. The feedback buttons allows users tell us how we’re doing. When users click 👍, Ravenna will ask if they need additional help.

When they click 👎, Ravenna will prompt the user to create a ticket by sending a Create Ticket button. Alternatively, if you would like Ravenna to auto create a ticket when 👎 button is clicked, you can configure via the following Queue setting.

When enabled, Ravenna will create the ticket and notify the user in the thread.

Both interactions are captured in Ravenna’s reporting, giving your teams valuable insights into the accuracy and effectiveness of Ravenna AI.

Request type classification

Teams often struggle to get employees to fill out the right Request type forms. With this feature you can allow your employees to converse with Ravenna AI agent in Slack as if they were talking to a human member of your team. Ravenna AI will decide based on the available knowledge if it can deflect the request with a knowledge answer, or if there is a request type that matches what they’re asking about. If both, it often will provide them with some information from the knowledge and offer to open a ticket for them.

Combining this feature with Answers in Slack allows your team to spend more time focused on responding to properly filled out requests and saves time from constantly answering the same questions.

Form-enabled vs. Form-disabled Request Types

When configuring request types for AI classification, you can control whether users fill out a form or tickets are created automatically using the Form enabled toggle in the request type’s AI settings.

Form Enabled (Default)

When a request type has forms enabled, Ravenna will present the configured form to users in Slack before creating the ticket. This ensures all required information is collected upfront.

Use this when:

• The request requires specific structured data
• Multiple custom fields need to be filled
• You want to ensure data quality and completeness

Configure the Form enabled setting in the request type’s Details tab under the AI section. This setting works seamlessly with AI classification to create the optimal ticket creation experience for each request type.

Auto tags

Ticket tags are a powerful way to create specific views and track specific data points via analytics.

Enabling AI tags means that whenever your employees open new tickets or add comments to existing tickets, Ravenna AI will automatically add tags when appropriate. This means your teams don’t have to.

---

Personality

Ravenna lets you tailor your AI agent’s personality to match your company’s voice and values. This creates a consistent, on-brand experience for users. Test your AI in realistic scenarios to fine-tune responses before going live.

Click the AI Personality tab along the top of the Knowledge page.

Use this page to define your AI’s voice, tone, and personality, so it communicates your brand effectively.

Custom Prompt

Set the core instructions that guide your AI agent’s behavior and communication style. This keeps responses consistent with your brand and goals.

Greeting Instructions

Define how your agent opens conversations — friendly, professional, or context-specific — to set the right tone.

Response Length Preference

Choose whether your agent’s answers are concise, moderately detailed, or comprehensive based on your audience’s needs.

Agent's Tone of Voice

Shape your agent’s personality: professional, casual, humorous, or anything in between.

Use of Emojis

Decide if and how your agent uses emojis to keep communication professional or make it more approachable and expressive.

After configuring, click Publish Changes in the top right corner.