Expert Chatbot Service for Telecommunication Services

Overview

Providing telecommunications information and services in a way that users can find conveniently and understand easily will establish the platform as a go-to service for everyone. To this end, the Korea Association of ICT Promotion (KAIT) has developed services such as real-time civil complaint consultation and personalized rate plan recommendations.

Currently, complaint handling is only available during weekday daylight hours and is limited to agents responding directly to user questions. This structure faces challenges in recommending necessary precautions or customized products for users selecting telecom services. Therefore, the project aims to introduce an interactive AI chatbot service to overcome these limitations.

Project description

The project developed a real-time specialized chatbot service for telecommunications by training on data such as legal regulations, terms of use, subscriber information (rate plans, usage patterns, early termination fees), and telecom service details.

By learning from civil complaints, terms of service, and legal data, the chatbot is capable of responding to inquiries in real-time at a professional level equivalent to a human agent. It analyzes words and sentences entered by users to provide natural answers and recommendations. Furthermore, it communicates with users in a conversational manner—following rational telecom consumption procedures—to recommend the most suitable rate plans tailored to their specific needs.

Key technological features

• Data Preprocessing: Preprocessed 310 data passages related to "Smart Choice" to optimize them for AI utilization
• Training & Indexing: Applied SearchGPT to the preprocessed data and implemented a chatbot solution featuring query generation based on SmartChat
• Hybrid Service Model: A hybrid system combining SearchGPT-based search/summarization with SmartChat-based chatbot services

Expected outcomes

Through the interactive AI chatbot consultation service, necessary information is provided to users at the right time and place, contributing to the operational efficiency of service personnel. Additionally, it is possible to develop a "telecom lifestyle check-up" service by analyzing the history of user usage patterns.

Furthermore, the service aims to contribute to reducing national telecommunications costs by promoting the refund of unclaimed telecommunications credits and providing guidance on discount systems. By implementing Hyper-scale AI in public complaint response services, the project will continue to drive improvements in overall service quality.

Launch year: 2023

Attribution

Case study courtesy of the National Information Society Agency, Republic of Korea, Apolitical's Content Partner. Sourced from Use Cases of Public AI Service, Vol.2, https://eng.nia.or.kr/site/nia_eng/ex/bbs/View.do?cbIdx=31975&bcIdx=29182&parentSeq=29198