380: Customer Service's AI Shift: Zendesk CTO Adrian McDermott on Deterministic AI and Context Engineering

The integration of generative AI into customer service and software development serves not as a catalyst for mass job elimination, but as a mechanism to address systemic service debt and unlock human potential. Adrian McDermott, CTO of Zendesk, emphasizes that while AI automates repetitive tasks, it simultaneously empowers agents to handle complex inquiries through real-time, context-aware assistance. By transitioning from non-deterministic generative models to deterministic, guardrail-protected workflows, organizations can achieve consistent, high-quality outcomes. This evolution shifts the role of service professionals toward higher-value, empathetic problem-solving. Ultimately, the future of work relies on "context engineering"—the ability to continuously feed relevant institutional knowledge into AI systems—to create a persistent, learning-based partnership between humans and technology that enhances productivity and deepens customer connections.