From Apps to Intent: Ping An’s AI Connects Finance and Healthcare for 250 Million Customers

Ping An, China’s financial services and healthcare giant, is known for its sprawling digital ecosystem—with more than 10 apps covering insurance, banking, health and other services. As the ecosystem grew, customer journeys became increasingly fragmented: users had to manage multiple logins, complex navigation and confusing jargon. To address this complexity, Ping An set out to rethink the user experience.

In a recent interview, Dr. Ni, Healthcare Scientist at Ping An’s technology unit, and Eileen Tu, Product Lead for AI Express Service, explained how Ping An’s new agentic AI innovation—Express Service—is driving a shift from multiple app-based customer interfaces to a single intent-based interface.

One Conversation: From “Apps as Interface” to “Intent as Interface”

Rather than launching yet another app, Ping An now prompts users to express their needs in simple, everyday language. This occurs through Ping An’s AI Express Service, which provides an integrated entry point to services across its app portfolio. The powerful AI agents then coordinate everything in the background across banking, insurance, healthcare and even senior care, removing friction that previously frustrated customers.

Tu describes the transformation: “Moving to ‘intent as interface’ means the customer simply clicks one button on our app and says in their own words what they want to do. The AI assistant handles the intent in the background across banking, insurance, and healthcare.” What was once a fragmented, multi-step journey now starts with a single conversation—technology dissolves barriers, so users can get things done with unprecedented ease.

Part 1: Building the Technical Moat: Inside Express Service’s AI Orchestration

Beneath the surface of this simplicity for the customer, there is sophisticated technical coordination. Dr. Ni explains that Express Service covers more than 300 digital services. Even a straightforward customer request triggers a multi-step automated process.

First, AI analyzes and recognizes the user’s intention. Next, it verifies the customer’s identity and access permissions, a critical safeguard in financial and healthcare services. Third is cross-system orchestration: many requests, such as a critical illness claim, require pulling data from policy management, claims processing, medical records and payments. Fourth, compliance rules are checked at every stage to ensure regulatory boundaries are never crossed. Finally, the agents not only provide answers but proactively advance workflows—helping upload documents if needed or routing cases to human specialists as needed.

Traditionally, these processes would require a customer to use several apps or make multiple phone calls. Express Service processes the requirements in seconds.

Part 2: The Economic Payoff: How Agentic AI Drives Sales and Retention

Agentic AI is not just an interface upgrade, it is driving measurable business results. Serving existing customers across Ping An’s ecosystem is 30 to 45% more efficient than acquiring new customers.

Agentic AI boosts both sales and customer stickiness. In 2025, Ping An’s AI agents delivered over RMB 133 billion (USD $19.64 billion) in sales through demand analysis and personalized recommendations. For the healthcare insurance business, AI-powered migration within Ping An’s platform generates higher-ticket policy sales at lower cost than external channels. Also, customers who interact with the service log in twice as often and demonstrate higher three-month retention rates.

Part 3: Why General AI Models Alone Can’t Compete in Finance and Healthcare

As generative AI and large language models (LLMs) proliferate across global finance, scale alone is not enough for true competitive advantage. Dr. Ni points to Ping An’s real-world, closed-loop ecosystem: “In financial services and healthcare, the real challenge is not building a larger model, but how to embed AI safely into professional workflows.”

Ping An’s AI agents operate across insurance, finance, and healthcare. It requires understanding of insurance claims, risk control, medical guidelines, disease management, and senior care. Decades of enterprise infrastructure and operational feedback give Ping An an edge that newcomers cannot simply buy or duplicate. As Dr. Ni puts it, people can buy the technology, but they cannot buy the ecosystem.

Part 4: Ethics, Risk, and Human Oversight in Agentic AI

What truly separates Express Service from traditional generative AI is accountability, Dr. Ni says. When an AI can take action—not just give answers—safeguards must be stringent. Dr. Ni describes four layers of governance in the system:

• Identity and permissions checks: Every action starts with rigorous identity and access verification. The AI agents can only operate within the scope of permissions granted to each user.

• Strict system access controls: Agents never interact directly with core systems. Instead, they trigger only pre-approved backend functions, each with strict guardrails in place.

• Human oversight for high-risk tasks: For higher-risk actions, a human-in-the-loop protocol ensures that critical decisions are reviewed and approved by qualified professionals before execution.

• Comprehensive audit trails: very tool used and every action taken by the agents is recorded. This allows the entire process to be traced for transparency and accountability.

In moving from app-centric to intent-driven services, Ping An is positioning AI not merely as a tool for efficiency, but as the front-end of a deeper, integrated, and secure user experience where accountability, compliance and customer trust are built in.

As financial players worldwide race to leverage AI, Ping An’s Express Service offers a potential roadmap: align advanced models with real-world processes and responsibilities, and put user intent—not app design—at the center of digital innovation.

Watch or listen to the full episode here.