Tower's implementation of an AI-enabled contact centre built on Amazon Connect has delivered measurable efficiency gains across its customer service operations, reducing average handling time by 2 minutes 38 seconds per interaction and cutting total customer call time by 796,000 minutes over seven months. The platform integrates real-time transcription, automated quality monitoring, and AI-assisted agent support directly into daily workflows rather than deploying these as isolated tools. This architectural choice—embedding AI into the interaction itself rather than bolting it on—means agents no longer pause to document calls manually, reducing cognitive load and the risk of missing critical customer context. Tower's 15% reduction in average handling time represents a material shift in how the insurer operates at scale, and the company's selection as the only APAC participant in AWS's inaugural PartnerLed Customer Success pilot suggests the approach has broader applicability beyond insurance.
The implications for CX teams are twofold. First, this demonstrates that AI-driven efficiency gains need not come at the expense of service quality—Tower's framing explicitly positions faster resolution as a trust-building mechanism rather than a cost-cutting exercise, which matters for teams navigating internal resistance to automation. Second, the real-time transcription capability addresses a persistent operational friction point: agents can now focus entirely on listening and problem-solving rather than splitting attention between conversation and documentation. For teams already running platforms like Zendesk or Salesforce Service Cloud, the question becomes whether your current architecture allows similar embedding of AI capabilities, or whether you're still operating in a mode where agents must context-switch between interaction and system entry. Tower's results suggest that the difference between incremental and transformational efficiency gains lies in how deeply AI is woven into the agent experience itself.
The broader strategic signal here is that cloud-native contact centre platforms with native AI capabilities are becoming table stakes for competitive service delivery. Tower's continued investment in AI-enablement throughout FY26 indicates this is not a one-off pilot but a sustained operational priority. For smaller vendors and teams with legacy infrastructure, the risk is not that AI will replace agents—Tower's model still requires human judgment and care—but that organisations failing to integrate AI into their core workflows will find themselves unable to match the efficiency benchmarks their competitors are now establishing.
Tower’s AI-Enabled Contact Centre Cuts Customer Call Time By More Than 796,000 Minutes Scoop - New Zealand News