NH NongHyup Bank cuts wait and call handling times about 20% with KT AI contact center

Before the project, NH NongHyup Bank's AI could directly handle only 45 tasks without connecting a caller to a human agent, and its speech-to-text accuracy was around 90%. Customer information and consultation histories were also scattered rather than managed on a single platform.

KT built a next-generation AI contact center (AICC) for NH NongHyup Bank in a roughly 40 billion won ($29 million) upgrade project, converting the consultation systems of the bank's branches and mutual finance operations nationwide to an AI-based platform. KT linked an AI call bot to the bank's core operating systems so it can look up customer and account information and carry out actual transactions. KT also built an AICC operations portal that manages channels including video consultation, chat and email in an integrated way, and migrated the large volume of data the bank already held to the new platform.

The number of tasks the AI handles directly, without connecting to a human agent, expanded to 180 from 45, and speech-to-text accuracy rose to about 97% from around 90%. Customer wait times and call handling times were cut by about 20% compared with before, and the workload on human agents was reduced. The two companies plan to keep strengthening AI call bots and generative-AI consultation support services.

• Link the call bot to core operating systems so it can look up accounts and carry out real transactions, not just answer questions.
• Manage every consultation channel and its history on one platform so customer service stays continuous and consistent.
• Track the number of tasks the bot completes on its own and its speech recognition accuracy, since both improved alongside shorter handling times.