Australian technology leaders are moving beyond AI-driven automation of routine support queries into more sophisticated applications that reshape how customer experience integrates with product design and operational efficiency. MYOB's Solo product exemplifies this shift: rather than bolting AI onto existing support infrastructure, the company embedded service as a foundational feature by designing the product around AI capabilities from inception. The result is a Zendesk-powered chatbot resolving 90% of issues—three to five times better than industry standard—with less than 1% of users attempting to bypass automation to reach humans. This approach required rethinking the entire end-to-end experience, moving from traditional UX questions to service-first architecture, and critically, it enabled non-technical community managers to build custom integrations using Zendesk App Builder in under a week. Guzman y Gomez and Aware Super demonstrate how agentic AI extends beyond customer-facing channels into operational decision-making and member guidance. GYG's kitchen management system uses AI to dynamically allocate orders between preparation lines and forecast when to activate secondary capacity, reducing wait times whilst balancing crew workload—a use case that raises an important question: as these systems mature, how will support teams need to evolve to handle exceptions and edge cases that AI cannot resolve autonomously? Aware Super's agentic AI coach for financial literacy reveals an unexpected insight: some members prefer discussing finances with AI rather than humans, yet the organisation's competitive advantage lies in proprietary member data, regulatory status, and data governance—factors that free consumer tools cannot replicate. This suggests that CX teams should be evaluating not just whether to deploy agentic AI, but whether their organisation's data assets and trust position create defensible differentiation.
Zendesk's own evolution signals where the broader CX technology stack is heading. Beyond automating routine interactions, the platform is moving into supervisory roles—AI monitoring agent conversations for quality coaching—and preparatory work, where long-running agents analyse tickets, extract backend data, and compile notes before human agents engage. McDermott's observation that developers using agentic AI are now 50 times more productive than previous "10x" performers carries direct implications for CX operations: the productivity gap will widen dramatically between teams that effectively leverage AI agents and those that do not. The shift from task-focused to purpose-focused work—where contact centre workers move from "handling enquiries" to "improving satisfaction and attracting customers"—requires fundamentally different skill sets and governance structures. For Zendesk administrators and support leaders, this means the build-versus-buy decision for agentic AI is no longer purely technical; it is strategic. Organisations that delay investing in governance, data quality, and staff enablement risk falling behind competitors who have already embedded AI into their operational DNA. The Australian case studies reveal that success depends less on the sophistication of the AI model and more on whether the organisation has redesigned workflows, upskilled teams, and aligned incentives around the new capabilities—a challenge that extends far beyond the support function itself.
How Australian firms are using AI in customer experience Computer Weekly