Safely manage your Zendesk from the AI assistant you already use, via the Deltastring MCP. Beacon configuration platform
← Back to news

KT builds next-generation AICC for NH Nonghyup Bank

KT has completed a 40 billion won AICC implementation for NH Nonghyup Bank that demonstrates the maturation of AI-driven contact centre architecture in financial services. The project expanded autonomous AI task handling from 45 to 180 use cases, improved speech recognition accuracy to 97 percent, and integrated AI callbots directly with core banking systems to enable transaction processing without agent handoff. Critically, KT built a unified operations portal managing video, chat, and email channels alongside voice—a technical integration that most legacy contact centre platforms still struggle to deliver seamlessly. The implementation signals that financial institutions are moving beyond chatbots as novelty features toward embedding AI as the primary transaction layer, with human agents repositioned as exception handlers rather than first-line responders.

The strategic implications cut across two dimensions. First, this raises a pressing question for teams currently managing multi-channel platforms: if a single vendor can now orchestrate transaction-capable AI across voice, video, and asynchronous channels with 97 percent STT accuracy, what is the competitive moat of traditional AICC vendors that haven't achieved this level of system integration? KT's ability to link AI directly to core banking systems—enabling actual fund transfers and account modifications without escalation—represents a capability gap that will force platform vendors to either deepen their banking integrations or cede market share in regulated verticals. Second, the 135-use-case expansion in autonomous handling suggests that the real value isn't in AI quality alone but in operational redesign: teams will need to fundamentally rethink agent workflows, training, and performance metrics if AI is handling four times more scenarios than it did previously. This isn't a dial-turn upgrade; it's a structural shift in how contact centres allocate labour.

The broader pattern here aligns with similar moves in other financial verticals—Bajaj Finance's AI bots handled 71 percent of DIY service requests in Q1 FY27—suggesting that financial services is becoming the proving ground for autonomous contact centre models. For CX leaders, the question becomes whether to wait for platform vendors to catch up with KT's integration depth or to pursue custom implementations that risk vendor lock-in. The fact that KT and NH Nonghyup Bank are now planning further expansion into generative AI consultation support indicates that this isn't a completed project but a foundation for continuous capability escalation, meaning teams should expect the competitive baseline to shift significantly within 12–18 months.