Zendesk launches AI agents built to resolve customer issues end-to-end
Zendesk announced a suite of AI-powered service tools designed to handle complete customer interactions without human intervention. The company unveiled the Resolution Platform at its Relate conference, positioning autonomous agents as core infrastructure rather than supplementary automation.
The platform combines AI agents, copilots, and workflow automation into a single operational layer. Unlike traditional chatbots designed to deflect tickets, these agents are built to resolve issues across email, messaging, voice, and third-party systems like ChatGPT and Gemini.
How the system learns and improves
The Resolution Learning Loop trains on nearly 20 billion support interactions. It continuously analyzes customer conversations to identify gaps in AI responses and refine automation in real time, creating a self-improving system that evolves based on actual support data.
A Context Graph stores historical analyses and reasoning patterns, allowing the platform to make better recommendations in future interactions. This operational memory layer helps agents understand the full history of a customer's issue across multiple touchpoints.
Building custom agents without coding
Agent Builder lets support teams create custom AI agents tailored to specific business workflows and policies. Organizations can automate front-office, middle-office, and back-office tasks while maintaining centralized oversight and governance.
The platform supports voice AI agents that handle conversations in more than 60 languages and can switch languages mid-interaction without losing context. This addresses the reality of multinational support teams where customers may speak multiple languages.
Support for employees and cross-system integration
Zendesk extended AI agents to internal support, embedding them into Slack and Microsoft Teams. These agents can search enterprise systems while enforcing role-based permissions, providing employees with secure, context-aware assistance.
The company also introduced copilots for different roles: agents get workflow optimization support, administrators receive operational analytics, knowledge teams get management tools, and analysts access automated recommendations.
Support for the Model Context Protocol allows agents to connect with external enterprise systems, letting organizations expand capabilities as new AI tools become available.
A shift toward outcome-based pricing
Zendesk moved to outcome-based pricing, charging customers only for interactions that AI agents resolve end-to-end. This represents a shift away from usage-based models toward pricing tied to measurable business results.
For support professionals, the shift means clearer accountability for AI performance and a focus on actual issue resolution rather than deflection metrics.
Learn more about AI for Customer Support and AI Agents & Automation to understand how these tools fit into broader service operations.
Your membership also unlocks:
