Transforming "Passive to Active" AI Customer Service: Proactive Care Increases Sales Performance

With hundreds of millions of customer interaction data, NetTalk has been leading the Taiwan market for 7 years by utilizing machine learning and natural language processing (NLP) technologies to create customer service and food ordering guide robots. They have accumulated interactions reaching tens of millions and have successfully assisted Shin Kong Mitsukoshi, Eslite, Swire Coca-Cola, Electrolux, KFC, and other multinational brands in deploying omnichannel customer service robots. Chen said that recent applications of new LLM and machine learning technologies have significantly improved customer interaction experiences, "the speed exceeds the growth of the past two to three years."

"Leveraging AI to assist enterprises in digital transformation, providing personalized services to customers, increasing service capacity while ensuring high-quality human-computer interaction experiences" has been a greatly increased demand in customer inquiries for NetTalk this year. Chen further explained that NetTalk's latest AI model can help enterprises understand customer intent, solve problems, and accurately meet customer needs more efficiently, even providing personalized care and product recommendations in a proactive manner, transforming "passive to active". It can also increase service capacity by more than 50% while maintaining high service satisfaction.

Currently, NetTalk's Chatbots are mainly utilized in e-commerce, retail, health food, home appliances, mobility, finance, and other major industries. Combined with the latest AI technology, they will gradually launch service robots with the most comprehensive industry knowledge, caring robots that understand problems and greet customers warmly, and sales robots that proactively analyze customer needs and make purchase recommendations. The head of the AI R&D team, Qi Deyu, stated that "collaborating with AI can significantly improve the productivity of frontline service personnel, enhance service experiences, and actively engage customers to increase sales performance."

The NetTalk AI R&D team participated in the ROCLING 2023 conference and presented the championship paper

Note
ROCLING is an annual conference for computational linguistics and speech processing in Taiwan, sponsored by the Chinese Language Processing and Computational Linguistics Association (ACLCLP). This conference provides an international forum for researchers and industry practitioners to share new ideas, original research findings, and practical development experiences in the fields of all languages and speech research. The ROCLING conference was initiated by ACLCLP in 1988, with the main goal of providing a platform for researchers and professionals around the world to share their experiences in natural language processing and speech processing.

Efficiency Increased by 60 Times: NetTalk’s Champion AI Interactive Responses are More Accurate!

ROCLING 2023  is organized by the Association for Computational Linguistics of the Republic of China, and it is one of the few data competitions primarily conducted in Traditional Chinese. In this competition, the NetTalk R&D team utilized the latest AI model to achieve a high score of 69.55 in the named entity recognition (NER) task for medical terms prepared by the organizing committee, winning the championship in 2023.

Wang Yinxue, a member of the participating AI R&D team, shared that although the competition theme was limited to the medical field, it actually involved a wide range of knowledge from multiple professional backgrounds such as chemistry, human biology, physiology, and nutrition, making the challenges very high. Therefore, the team incorporated generative AI technology based on NetTalk AI, which greatly improved the efficiency of task recognition. Team member Wu Wenhong also emphasized that this innovative technology "allows AI to quickly grasp key information in limited data, providing more accurate and detailed answers."

Qi Deyu further explained that the latest AI model not only improves the accuracy of Chatbot answers but also increases the efficiency of training AI in automated system processes by 60 times, "not only speeding up the deployment of AI chatbots for enterprise customers but also allowing them to handle larger volumes of knowledge in a shorter time, reducing the cost for enterprises to build AI themselves."

Chen stated that NetTalk AI will continue to evolve and innovate, through three types of omnichannel robots for service, care, and sales, covering the entire customer journey of brand management, integrating knowledge from eight major industries, creating the warmest human-computer interactions, achieving personalized customer experiences, maximizing the value of private traffic, accelerating digital transformation of brand enterprises, and creating a competitive market advantage.

The NetTalk AI R&D team collaborated with Professor Zhang Zhixing from National Taiwan University’s Department of Computer Science to win the championship in the named entity recognition competition in the medical field at ROCLING 2023