Trusting Social revolutionizes customer experience in the financial sector with generative AI agents built on Azure

Traditional credit scoring methods often leave behind a significant portion of the population underbanked. Over the last 10 years, Trusting Social has worked to fix this issue and deliver financial access to all. The company leveraged advancing data science and cutting-edge technology to perfect its credit scoring models and tailor its customer acquisition engines for emerging markets and underbanked consumers.

“I believe that personalized banking is a right and not a privilege, which is the case today,” explains Nguyen Nguyen, Chief Executive Officer of Trusting Social. “With our platform, we are working to reduce the cost of banking so that everyone can access lending and banking services. Today, we serve over 1 billion subscribers across Asia.”

The global shift towards AI opened new opportunities for the company to expand its portfolio. “AI has the power to amplify productivity, which in turn will lead to prosperity, social mobility, and stability,” he says. To do this successfully, Trusting Social needed a like-minded partner to support its digital transformation and AI journey.

A comprehensive solution and solid support

With operations across Asia, Nguyen’s primary concern was finding a partner that could work seamlessly across borders while complying with regional policies and governmental regulations. “Microsoft was the obvious choice. It’s at the forefront of the global shift towards AI and offers a flexible and comprehensive platform to build what we need,” explains Nguyen.

The first step was laying the data foundation. Trusting Social integrated its data platform with Azure Kubernetes Services, Azure Database for PostgreSQL, and Azure Databricks to quickly start developing and deploying solutions. Azure Synapse Analytics helps bring insights across to the company’s data warehouse rapidly. Nguyen believes the transition happened seamlessly due to Azure's simplified migration process and compatibility with industry standards. “There’s a comprehensive online training platform that provides users with a wide range of lessons and rich content that’s extremely helpful. In addition to this, Microsoft also offers frequent workshops to share insights on optimizing businesses with Azure,” he says.

Trusting Social also relied on Microsoft to meet the stringent security requirements within the financial sector. This included setting up strict data governance protocols, utilizing Microsoft Sentinel for secure central logging, and installing Web Firewall Applications (WAF), and Intrusion Detection and Prevention Systems (IDS/IPS) for public-facing internet interactions. “Since the Azure platform offers a comprehensive range of tools, we don’t need to add on any third-party solutions. This not only ensures our data security, but also minimizes potential risks while offering us the flexibility to offer our clients specific security settings based on their requirements—whether it’s end user protection, product system security, or data leak prevention,” Nguyen explains.

Unparalleled productivity with AI

After six months of collaboration with Microsoft teams supporting digital natives and leveraging the Microsoft for Startups Founders Hub, Trusting Social successfully introduced an e-Know Your Customer (eKYC), which would play a critical role in efficiently validating the identity of unbanked individuals. As one of the earliest adopters of Azure OpenAI Service in Vietnam, the company also introduced the Agent Foundry platform—a generative AI platform, with multiple autonomous AI agents. “With Azure Machine Learning and Azure OpenAI Service, we’re able to build and train these generative AI agents with our own models and those available on the Azure AI platform to meet the specific needs of organizations,” explains Ha Le, Director of Product Innovation at Trusting Social.

By utilizing a combination of Big Data predictive modeling, vast industry domain knowledge, and many digital skills, each agent represents a significant leap in automating complex tasks and scaling operations. “Whether it’s using office applications or doing advanced analytics and coding, these agents can be trained to take over a multitude of tasks that will exponentially boost productivity and increase customer loyalty, product penetration, and eventually increase revenue per customer,” explains Le.

The first of these agents, ALICE—a sales and relationship management agent that can help brands reach out to millions of consumers—has already been launched in Vietnam. “ALICE delivers helpful and personalized care for customers throughout their banking journey, from onboarding to answering customer queries, and handling conversational cross-sales when needed. Since deploying ALICE, one of our clients has seen a 10% increase in conversion rates in its credit card business,” says Nguyen. “These results are incredible. With ALICE, there has been an increase in customer satisfaction (CSAT) by 42% and a 10% increase in spending for engaged customers.”

Two additional agents, ANANDA, a business and data analytics agent, and ALAN, a software agent, are also currently being developed. Nguyen believes that these trainable, AI-powered autonomous agents on the Agent Foundry platform will transform how organizations interact with customers. “With these agents, we can empower human workers to focus on strategic tasks and problem-solving rather than wasting time on tedious, menial tasks,” he says. 

Moving beyond the financial sector

Nguyen is confident that the best is yet to come. The next step for the company is expanding its services into other sectors, including consumer goods, telecommunications, and professional services. “Agent Foundry is just the beginning of a radical transformation in how businesses will operate in the region. One day in the future, we envision workers relying on multiple AI-driven agents to carry out daily tasks. This will amplify productivity and reduce the cost of servicing mass consumers drastically,” he concludes.