FPT.AI Virtual Agent for Call center allows financial institutions to make tens of thousands of calls every day, fulfilling more than 90% of customers’ loan and service requests.
Digital transformation – The recent trend for optimal operation
Home Credit is a mass market consumer finance provider established in 1997 in the Czech Republic. Home Credit Vietnam officially operated in Vietnam since 2008; and currently has become one of leading consumer finance companies in Vietnam market with a huge number of customers and transaction points nationwide.
With more than 11 million customers in Vietnam, Home Credit maintains a customer care team nationwide, especially operators who solve all complaints or inquiries of customer, answer questions and remind payment deadline. On average, the customer care team has to handle thousands of incoming and outgoing calls every day.
In recent years, Home Credit has implemented its digital transformation roadmap. Specifically, the company has started to adopt new, emerging technologies, such as Big data and AI. In addition, it has cooperated with FPT to launch FPT.AI Virtual Agent for Call center, the first Virtual Agent in Vietnam, to build a smart customer service system.
According to Nguyen Thuong Tuong Minh, Project manager of FPT.AI, financial institutions put faith in Virtual Agent for customer experience improvement and optimization of employee performance.
Previously, in China, Home Credit implemented digital lending and fully processed about 96% of loans. Furthermore, Home Credit also invested in virtual assistant (chatbot/voicebot) to handle 200,000 calls a day in China and solve up to 80% of repetitive tasks at Home Credit India.
Human – Machine collaboration
According to FPT’s representative, the biggest challenge facing FPT is to digitize communication with customers over the phone by capturing customer behavior when they use consumer finance services, and design flexible and correct call scripts for Virtual Agent.
After the process of researching, designing and adjusting scripts for phone calls, FPT and Home Credit have created two scripts for the automated virtual operator to serve two different needs of hundreds of thousands of customers, which are to make Pre-collection call and Welcome call.
Firstly, Pre-collection call is to notify customers the payment deadline. FPT.AI Virtual Agent for Call center makes business calls to remind customers the loan due date and collects customer information, including time, payment method and payment confirmation, through two-way automated customer service interactions. Then, the scripts will automatically adjust to the information and collate it to make recommendations to customers.
Secondly, in term of Welcome call, FPT.AI Virtual Agent automatically makes calls to first-time customers to verify customer information and introduce products and services; and forward the information to management system.
“As Vietnamese is unique and diverse, we have tested and piloted various complicated scripts to identify exact customers’ intent in the call center environment, especially in the noisy environment,” said FPT.AI representative.
What makes FPT.AI Virtual Agent stand out from traditional call center service is the ability to understand customers’ intents and extract meaningful information for each intent, helping it give correct answer and keep a friendly and seamless conversation.
“Customers want to grasp such information as who is calling or how they can get support within just a few seconds,” said Minh. He added that the welcome call is solely to reach new customers, while the pre-collection call is a challenge for FPT.AI, as it requires flexible conversation scripts that can adjust to information gathered from customer’s answer.
To apply these features to the Virtual Agent services of financial institutions, natural language processing (NLP) is integrated into FPT.AI Virtual Agent. This technology includes acoustic and linguistic models which are optimized for communications over phone calls. The models can become “intelligent” through analysis and training process based on collected data. The conversation between virtual agent and customer is stored in the system in two formats, which are text and audio in real time, allowing administrators to listen to the conversation or extract necessary information.
“At the same time, administrators in financial institutions can monitor online conversations via Livechat interface between virtual agent and customer,” said FPT.AI representative. After 6 months of operation, FPT.AI Virtual Agent can make tens of thousands of automated conversations a day, fulfilling more than 90% of customers’ loan and service requests.
FPT.AI representative emphasized that although FPT.AI Virtual Agent has just been put into operation for 6 months, the success rate of calls to date has reached over 90%. That means 9 out of 10 calls will be hung up after the virtual agent completes customer’s request. This achievement, said Minh, comes from continuous improvements that allow the virtual agent to learn better and come closer to customer habits. Every week, FPT.AI updates new features and adjusts its weaknesses.
Branislav Vargic, Chief Operations Officer of Home Credit Vietnam, said that the smart Virtual Agent for Call Center is helping the company reach and serve a large number of customers at the same time while improving service quality and customer experience. “This is an important milestone for Home Credit in particular and the consumer finance industry in general,” said the COO of consumer loan company.
According to FPT’s representative, thanks to deployment on the cloud platform, FPT.AI Virtual Agent for Call Center can be scalable to handle thousands of customer requests at the same time without facing overwhelming call centers. The solution can be integrated into businesses’ call centers or cooperated with other call centers easily via APIs.
In the new normal, automated call centers combined with Artificial Intelligence will soon show great potential in optimizing customer service and improving productivity for businesses. “It promises to be an indispensable customer care tool in the industry 4.0 era, especially for the banking and financial service industry”, said FPT representative.