Skip to main content

AI Assistant

PurposeThe AI Assistant allows users to explore and interact with their Raven team’s data using generative AI, enabling seamless analysis, actionable insights, and secure workflows tailored to organizational needs.
Last UpdatedMarch 26, 2025

What is Raven's AI Assistant?

The AI Assistant is a transformative tool that empowers organizations to explore and interact with data integrated within their Raven team. By leveraging generative AI chat functions powered by industry-leading Large Language Models (LLMs), the AI Assistant enables users to ask questions, analyze data, and perform actions—all within the context of their team’s alerts, data points, and integrations.

Unlike standalone AI tools, the AI Assistant operates seamlessly within the Raven platform, ensuring every interaction is tailored to your organization’s unique datasets and workflows. Designed for simplicity and security, it bridges the gap between complex datasets and actionable insights through intuitive text-based conversations.

How to Access the AI Assistant​

  1. From Sidebar Navigation

    • Click on the AI Assistant App in the sidebar menu.
    • The assistant will open in full-screen view, allowing you to start a new thread or explore your data.

  2. From Quick Access Button

    • Click on the assistant button pinned at the bottom-right corner of the platform.
    • The assistant will open in mini view, enabling quick interactions while navigating other Raven apps.

Supported LLMs​

The AI Assistant supports a wide range of LLMs, ensuring flexibility based on your organization’s requirements:

Private Cloud LLMs (via AWS Bedrock):​

  • All Claude models
  • All Llama models
  • All Mistral models

Public Cloud LLMs:​

  • ChatGPT models
  • Google Gemini Pro

Assistants​

Depending on your team’s needs, multiple assistants can be configured. Each assistant is tailored with pre-configured prompts and specialized functions to handle specific tasks.

Assistant functions are assigned to each assistant and enable them to perfom actions such as:

  • Compiling and running Raven Queries
  • Sending emails
  • Creating visualizations
  • Generating PDFs
  • Creating alerts
  • Browsing URLs
  • Searching DuckDuckGo
  • Sending Slack messages
  • Accessing threat intelligence data
  • Updating alerts
  • Creating cases
  • Analyzing security events

DTACT’s engineering team can further customize assistant prompts and functions to align with your organization’s unique requirements.

Threads​

Each conversation with the AI Raven’s AI Assistant represents a thread. Threads are allow users to maintain context across conversations with the AI Assistant.

The following options are available for managing threads:

  1. Thread Visibility

    • Set visibility to:

      • Private: Only visible to you.
      • Team: Shared with your entire team.

  2. Thread Actions (via Three-Dot Menu)

    • Delete a thread.
    • Start a new thread.
    • Pin a thread for easy access in mini view.
    • Edit thread details (name, summary, schedule, LLM, visibility).
    • Fork a thread to restart a conversation while keeping its context.

  3. Mini Assistant Mode

    • Switch to mini assistant mode to keep the thread accessible in a smaller screen while browsing other Raven apps.

  4. Threads History

    • The Threads History feature allows users to revisit and manage past conversations with the AI Assistant:

      • View all previous private or team threads.
      • Search for specific threads by name or content.
      • Fork a previous thread to start a new conversation while feeding it with prior context.

Best Practices​

  1. Use descriptive names for threads to make them easily identifiable.
  2. Pin important threads in mini view for quick access during multitasking.
  3. Regularly review Threads History to leverage prior discussions for new insights.