Zendesk AI Agents

Knowledge-grounded player-support conversations with configured procedures, selected system actions and human escalation.

Conversational iGaming assistants

Compare product Official product page

The offering

What Zendesk AI Agents does

Zendesk AI Agents connect knowledge sources to conversations and configured procedures. The current iGaming & Betting page explicitly markets the product to gambling operators. Current setup documentation covers messaging, email and voice agents, with each agent assigned one channel type, a brand/knowledge scope, languages and system replies.

Define the player requests the assistant can answer and the sensitive cases it should hand to staff. Keep authenticated account data and backend actions behind agreed permissions, with separate checks for completion and failure. Test knowledge gaps and escalation before activation. The Knowledge reply can branch to an answer or fallback, and Zendesk documents an immediate messaging handoff that collects customer information. Automated-resolution allowances and enabled functionality depend on plan.

Who should explore it?

Gambling support teams deploying scoped conversational assistance with controlled data access and human fallback.

Answer an in-scope player help request from approved knowledge
Collect support context and route a sensitive or unresolved request to staff

Documented capabilities

Connected knowledge answers
Configured conversational procedures
Separate channel-type agents
Language and system-reply setup
Human escalation workflows

Reported in the linked public sources. We have not independently tested these capabilities.

Product details

Request scope: Configured support use cases and knowledge answers; gambling-specific account or payment actions require their own permitted workflow.Source
Deployment channels: Current creation guide covers messaging, email and voice; each AI agent operates on one channel type.Source
Knowledge context: Selected brand/help center and connected external sources supply knowledge. Authenticated player-record access needs explicit integration scope.Source
Action boundary: Procedures can use connected tools/systems; configure permissions and validate completion/failure for each operator action.Source
Human handoff: Documented messaging flow can collect customer information and immediately escalate to a human agent.Source
Setup and fallback: Configure and test system replies; knowledge-found/no-knowledge branches and escalation can be customized.Source
Usage model: Automated resolutions with plan-based allowance; confirm the contracted functionality and overage scope.Source

Delivery & commercial details

Delivery & integration: Configure knowledge, brand, channel, languages and replies in the AI agents workspace; test before activation. Scope backend tools and authentication separately from the knowledge connection.Source
Pricing: Usage is measured through automated resolutions, with an included allowance based on plan. Confirm enabled agent functionality, channel setup and additional usage in the proposal.Source

Confirm market availability, rights, service levels and contractual terms for your implementation. API availability alone does not establish a named integration.

Buyer questions, answered

What happens when the assistant cannot find a knowledge answer?

The documented Knowledge reply branches on whether knowledge is found. No-answer cases use a default or escalation reply, and the procedure can be configured to escalate after repeated failed searches.

Source

Can we collect context and immediately hand the conversation to staff?

Yes. Zendesk documents a messaging workflow that collects customer information and immediately escalates to a human agent. Test the collected fields, queue assignment and failure route.