Transforming contact center operations on AWS with Tower Insurance

Tower has been protecting New Zealand families and businesses for more than 150 years, offering coverage across home, contents, car, and commercial lines. With a customer base of approximately 300,000 policyholders, the company depends on its contact center to deliver responsive, personalized service at scale. Tower is one of New Zealand’s longest-established insurers, and the company’s reputation for service quality is central to its competitive position. That mission became increasingly difficult as Tower’s systems grew more fragmented over time and customer expectations for fast, seamless service continued to rise.

Human agents juggled disconnected tools and manual workflows, creating what Tower’s team described as “swivel-chair” activity—constantly switching between applications to serve a single customer. Post-call administrative work consumed significant agent time, with every completed call requiring manual wrap-up before the agent could move on. Quality assurance teams could review only a fraction of interactions, leaving large gaps in visibility into service performance across the organization. Without automated monitoring, coaching opportunities were missed and service inconsistencies went unaddressed. These inefficiencies rippled outward: Customers experienced longer wait times, agents carried higher workloads and stress, and management lacked the data needed to staff and coach effectively. Tower recognized that incremental improvements would not be enough. The company needed a foundational change to how its contact center operated.

Tower selected AWS and Deloitte for their shared commitment to rapid, responsible innovation at enterprise scale. Deloitte brought deep contact center transformation expertise and a proven delivery record of accelerating adoption; AWS provided a secure, cloud‑based foundation for Tower to embed AI directly into customer operations. Rather than introducing AI through isolated pilots, Deloitte took Tower through the BVR program with a focused 10‑week AI sprint. This supported rapid activation of high‑value AI capabilities while sustaining alignment with enterprise architecture, governance, and New Zealand’s regulatory requirements. Together, the three organizations operated as a single team, moving from pilot to production in 10 weeks.

The solution, built on Amazon Connect, is a single, unified agent workspace that eliminates swivel‑chair activity across disconnected tools. Using virtually unlimited AI on Amazon Connect, Tower implemented call transcription and summarization to automate post‑call work and improve agent productivity. The platform was rapidly extended to include email and campaign capabilities, supporting proactive customer engagement at scale.

AI was embedded across the contact center value chain. Frontline agents received near real‑time assistance through Amazon Connect AI agents, transforming customer service with agentic AI by reducing reliance on subject‑matter experts and accelerating resolution times. Managers and quality teams used Amazon Connect Contact Lens and generative AI–driven evaluation forms to automate performance assessments for near real-time analytics, significantly expanding quality assurance coverage and insight. Tower also adopted Amazon Bedrock—a service for building generative AI applications and agents—establishing a secure foundation for future AI use cases, including customer self‑service assistants and advanced analytics. Together, these capabilities empowered Tower to operationalize AI not just as a technology, but as a core part of day‑to‑day customer service delivery.

Deloitte’s role extended beyond technology delivery. The team worked to evolve the Tower team’s mindsets and practices to sustain innovation after the solution was deployed. Within 10 weeks, Deloitte and Tower delivered five AI-driven use cases, which included sending proactive customer campaigns and extending Amazon Connect AI agents to sales teams.

Across the broader transformation and AI sprint, Tower onboarded 630 agents across 11 lines of business, deployed Amazon Connect with AI-powered call transcription and summarization, and rapidly expanded into additional use cases. Since the solution’s launch, Tower has reported a 26 percent reduction in average call handling time and an 18 percent reduction in average email handling time. At the same time, coverage of automated performance evaluations rose from 4 percent to 30 percent. With approximately 60 percent employee adoption of AI-driven capabilities, the new solution has helped Tower’s NPS increase 4 points and resulted in 492,000 customer minutes saved and 8,200 agent hours freed up—all while maintaining service stability.

“This has been a powerful proof point of what’s possible when AI ambition is matched with disciplined delivery,” says Kylie Bryant, partner at Deloitte. “In just 10 weeks, Tower demonstrated that AI-driven experiences can be delivered at pace, with stability, strong governance, and real business impact. This pilot has laid a credible foundation for scaling AI safely into business as usual, and it reflects the strength of the one-team partnership between Tower, Deloitte, and AWS.”