Introduction: Hangzhou Bank's customer service center uses intelligent technology to change the traditional manual operation mode and improve the level of intelligent operation and management, which has become an important breakthrough for the customer service center to accumulate data capabilities and realize business agile empowerment.
In recent years, the service scale of Hangzhou Bank's customer service center has continued to expand. Against the background of increasingly large data and increasingly complex business, intelligent technology is used to change the traditional manual operation method and improve the level of intelligent operation management to become the customer service center to accumulate data capabilities and achieve business agility. An important breakthrough in empowerment.
In order to strengthen the intelligent operation and continuously improve the product innovation ability, Alibaba Cloud Intelligent Customer Service relies on the leading AI technology of DAMO Academy to help Hangzhou Bank Customer Service Center to jointly build a one-stop full-link intelligent dialogue analysis platform, based on intelligent quality inspection In terms of capabilities, it deeply integrates actual business scenarios for data mining and insight analysis, and achieves dual-engine driving of precise service quality control and business value mining by accumulating massive unstructured data.
Engine 1: Accurately control the service quality of bank agents and consolidate refined management capabilities
Intelligent quality inspection replaces manual quality inspection to achieve full data coverage, visualization of service quality, customized data dashboards, and automatic accumulation and mining of data "gold mines", providing a solid foundation for the refinement of management capabilities. The intelligent dialogue analysis platform jointly built by Bank of Hangzhou and Alibaba Cloud Intelligent Customer Service has a complete intelligent quality inspection base, and provides five types of quality inspection operators including text, voice, knowledge, process and AI model . The quality inspection content covers the four dimensions of service specification, service attitude, service process, and emotion recognition . The rule engine can be flexibly configured according to the individual needs of the business, which further improves the accuracy of quality inspection, and the accuracy rate exceeds 90% in some scenarios . Under the uninterrupted operation of the platform 7*24 hours, through data transmission-voice structuring-quality inspection scheme matching-model rolling scan-quality inspection result generation, Hangzhou Bank Customer Service Center has realized 100% automatic quality inspection of dialogue data, and the quality inspection capacity It is nearly 30 times higher than before. From manual to AI, from sampling to full quantity, intelligent quality inspection effectively bridges the shortcomings of traditional quality inspection, such as poor timeliness, low coverage, and strong subjectivity. Relying on the visual data kanban, managers can see the service quality in an all-round way, and trainers can take targeted measures to help agents make up for shortcomings, plug loopholes, and strengths and weaknesses, and form a closed-loop service quality management system of "discovery-optimization-implementation" , continue to improve customer service experience.
Engine 2: Dig deep into the " gold mine " of data, gain insight into customer opportunities, and help mine business value
Powerful intelligent dialogue analysis capabilities help Hangzhou Bank's customer service center to transform into a smart operation center, so as to dynamically listen to customer needs, explore business opportunities in real time, and assist in the optimization of operation strategies in a timely manner. The intelligent dialogue analysis platform has a solid dialogue analysis model and excellent quality. It can carry out multi-dimensional analysis of a large amount of dialogue data generated by voice transcription, realize deep drilling of conversation content and customer voice insight, and extract data value information. On May 15, in the hot word cloud map of the intelligent dialogue analysis platform, the word "mortgage loan" was listed, and it rushed to the top 10 hot words. Relying on the data dashboard, operators can quickly locate customers to consult hot issues after viewing trend analysis of hot words and conversation details. It turned out that, in addition to the conventional mortgage repayment consultation, the People's Bank of China and the China Banking and Insurance Regulatory Commission issued a notice on the same day to adjust the lower limit of the interest rate of the first residential commercial personal housing loan, which led to more customers calling to inquire whether the mortgage interest rate was lowered. In response to the new policy on mortgage interest rates, operators have added relevant answers to the announcement platform of the customer service center in a timely manner to optimize service strategies.
The focus of hot business is due to the intelligent dialogue analysis team of Hangzhou Bank Customer Service Center, which has accumulated rich business prototypes with years of experience and innovative thinking. First, build a "7+4" two-way emotional hierarchical recognition system to achieve 11 detailed emotions. Sensitive perception, the second is to build a 3-level business label system, covering 2982 business labels, combined with the advantages of Alibaba DAMO Academy's algorithm capabilities, to achieve the transferability of algorithm solutions in different business scenarios, based on data labeling and optimization strategies, promote The platform continues to iteratively upgrade. The cooperation between the two parties has implemented the DialogCSE algorithm for dialogue semantic sentence vector encoding pioneered by Alibaba Dharma Academy , which can build a common dialogue semantic representation base based on the dialogue scenes of various businesses in the industry, and build rich business analysis on the base. Function, to achieve high accuracy, high recall and low operating costs on the algorithm model , reaching the industry-leading level.
Figure: Core Algorithm Model of Intelligent Dialogue Analysis
At present, the intelligent dialogue analysis platform mainly covers the following three scenarios:
1. Repeated call analysis: automatically identify the call data of the call center's daily repeated calls, monitor key indicators such as repeat call rate and one-time resolution rate in real time, and gain insight into repeat call intentions and repeat call links. The product can assist the business to locate the intentions and problems of high-frequency repeated calls, and ultimately reduce the rate of repeated calls by optimizing the quality of customer service.
2. Basic analysis: Basic analysis can present the specific call content distribution of the call center, and analyze the correlation and path between content topics, and clearly gain insight into what customers are saying every day through a report. what relationship". Help managers gain insight into and analyze the main content of customer conversations to capture customer voices and potential business opportunities.
3. Sentiment analysis: It can accurately identify the emotional intentions that need to be paid attention to on the customer side and the customer service side, such as public opinion crisis, problem escalation, complaints, insults, etc. on the customer side, and guidance complaints, rhetorical questions, evasion, etc. on the customer service side. In the service, real-time early warning can be used to remind customers to intervene in time for manual intervention in response to abnormal emotions of customers to reduce hidden risks. After the service, special training on emotion management and service attitude can be carried out to strengthen service awareness and improve customer satisfaction.
Through basic analysis of call content, perspective of repeated call links, dynamic identification and early warning of public opinion, and hot business focus, the analysis efficiency of the intelligent dialogue analysis platform can reach the minute level, which can efficiently assist operation managers to adjust operation strategies and promote customer service. The center plays a greater role in product optimization, process improvement and business expansion of the whole bank.
In the future, Alibaba Cloud's intelligent customer service team will continue to maintain in-depth cooperation with the customer service center of Bank of Hangzhou, and implement more new scenarios such as automatic generation of call content summaries, custom in-depth analysis of business scenarios, and customer portrait tag mining, helping the customer service center of Bank of Hangzhou to use data Drive innovation, improve data creativity, further build the customer service department into a "value center", and create and deepen a new business card for smart operations.
Copyright statement: The content of this article is contributed by Alibaba Cloud's real-name registered users. The copyright belongs to the original author. The Alibaba Cloud developer community does not own the copyright and does not assume the corresponding legal responsibility. For specific rules, please refer to the "Alibaba Cloud Developer Community User Service Agreement" and "Alibaba Cloud Developer Community Intellectual Property Protection Guidelines". If you find any content suspected of plagiarism in this community, fill out the infringement complaint form to report it. Once verified, this community will delete the allegedly infringing content immediately.
**粗体** _斜体_ [链接](http://example.com) `代码` - 列表 > 引用。你还可以使用@来通知其他用户。