For younger generations, paper bills, loan forms and even cash might as well be in a museum. Smartphones in hand, their financial services largely take place online.
The fintech companies that serve them are in a race to develop AI that can make sense of the vast amount of data companies collect, both to provide better customer service and to improve their own backend operations.
Fintech based in Vietnam MoMo has developed a super app that includes payment and financial transaction processing in a self-contained e-commerce platform. The convenience of this all-in-one mobile platform has already attracted more than 30 million users in Vietnam.
To improve the efficiency of the platform’s chatbots, know-your-customer (eKYC) systems, and recommendation engines, MoMo uses NVIDIA GPUs running in Google Cloud. It uses NVIDIA DGX systems for training and batch processing.
In just a few months, MoMo has achieved impressive results by accelerating the development of solutions that are more robust and easy to scale. Using NVIDIA GPUs for eKYC inference tasks resulted in a 10x speedup over CPU usage, according to the company. For the MoMo Face Payment service, using TensorRT reduced training and inference time by 10 times.
AI offers a different perspective
Tuan Trinh, director of data science at MoMo, describes his company’s use of AI as a way to gain a different perspective on its business. One such project processes large amounts of data and turns it into computerized visuals or graphics that can then be analyzed to improve connectivity between application users.
MoMo has developed its own AI algorithm that uses over a billion data points to drive recommendations for additional services and products to its customers. These offers help maintain a line of communication with the company’s user base, which helps drive engagement and conversion.
The company also deploys a recommendation box on the home screen of its super app. This has led to a dramatic improvement in its click-through rate as the AI provides customers with helpful recommendations and keeps them engaged.
With AI, MoMo says it can process the habits of 10 million active users over the last 30 to 60 days to train its predictive models. In addition, NVIDIA Triton Inference Server helps unify service flows for recommendation engines, greatly reducing the effort of deploying AI applications in production environments. In addition, TensorRT Helped triple the performance of MoMo’s Payment Services AI model inference, improving customer experience.
Chatbots move the conversation forward
MoMo will use AI-powered chatbots to allow it to scale faster when onboarding and engaging with users. Chatbot services are particularly effective on mobile device apps, which tend to be popular with younger users, who often prefer them to phone calls to customer service.
Chatbot users can learn about a product and get the support they need to evaluate it before buying it – all from a single interface – which is essential for a super app like this from MoMo which functions as a one-stop-shop.
Chatbots are also an effective vehicle for upselling or suggesting additional services, says MoMo. When combined with machine learning, it is possible to categorize target audiences for different products or services to personalize their experience with the app.
AI chatbots have the added benefit of freeing up MoMo’s customer service team to handle other important tasks.
Better credit score
Credit history data from all of MoMo’s more than 30 million users can be applied to models used for financial services risk control using AI algorithms. MoMo has applied credit scoring to lending services within its super app. Since the company doesn’t rely solely on traditional deep learning for less complex tasks, MoMo’s development team was able to achieve higher accuracy with faster processing times.
The MoMo app takes less than 2 seconds to make a loan decision, but is still able to reduce risky loan targets with more accurate AI predictions. This helps prevent customers from taking on too much debt and helps MoMo not miss out on potential revenue.
Since AI is able to process both structured and unstructured data, it is able to incorporate information beyond traditional credit scores, such as whether customers are spending their money on necessities. or luxury, in order to more accurately assess a borrower’s risk.
The Future of AI in Fintech
As fintechs increasingly apply AI to their massive data stores, the MoMo team predicts the industry will need to assess how to do so in a way that protects user data – or risk losing user loyalty. clients. MoMo already plans to expand its use of graphical neural networks and models based on its proven ability to dramatically improve operations.
The MoMo team also thinks that AI could one day make credit scores obsolete. Since AI is able to make decisions based on broader unstructured data, it is possible to determine loan approval by considering other risks besides a credit score. This would help open up the pool of potential users on fintech apps like MoMo to people in underserved and underbanked communities, who may not have credit scores, let alone “good”.
With about one in four “Underbanked” American adults, making it harder for them to get a loan or credit card, and more than half of the African population completely”invisible creditwhich refers to people without a bank or credit score, MoMo believes AI could bring banking access to communities like these and open up a new user base for fintech apps at the same time.