Skip to main content
RAUM AI is a customer-support AI platform with two product surfaces: AI Assistant for Zendesk teams and AI Chatbot for customer-facing website support. Both products can use your organization’s AI-provider configuration, procedures, guidelines, knowledge, and compatible connections, while keeping internal Zendesk context and public Chatbot access behind separate product boundaries.

Onboarding

Choose AI Assistant, AI Chatbot, or both, create your organization, connect an AI provider, and complete the setup required for your selected product.

AI Assistant

Configure Zendesk-assisted and autonomous support with procedures, knowledge, connections, quality controls, and operational analytics.

AI Chatbot

Build a public website widget with approved knowledge, visitor identity, forms and handoffs, public connections, topics, simulation, and analytics.

Zendesk App

Install and use the RAUM AI Assistant Marketplace app inside the Zendesk ticket editor and Ticket Insights sidebar.

Choose the right starting point

New organization

Start with Onboarding. The flow changes according to whether you select AI Assistant, AI Chatbot, or the bundle.

Zendesk support team

Use AI Assistant for agent-reviewed Copilot workflows or configured Agent Mode automation. The Zendesk App section covers the controls agents use inside Zendesk.

Website support

Use AI Chatbot for the customer-facing web widget. Public resources, live data access, visitor identity, and state-changing actions use a separate approval and safety boundary from internal Assistant resources.

Shared foundations

Across the platform, RAUM can use:
  • bring-your-own AI-provider credentials;
  • generation and embedding models;
  • Procedures;
  • Response AI Guidelines;
  • documents and websites;
  • reusable assets and temporary Updates; and
  • compatible API/MCP connections.
Product-specific rules determine which of those resources are available to an internal Zendesk agent versus a public website visitor.

Security and verification

RAUM’s current production architecture is hosted on Microsoft Azure. Sensitive stored credentials are encrypted by the application, and customer AI requests use the provider configuration attached to the organization. For detailed security and data-flow information, see Security overview. For public Chatbot controls, see Safety & Simulation.