Episode 1 – From Interface to Intent: How Agentic AI Reshapes the Economics of Finance and Healthcare
What if you could get things done with just one sentence?
Over the past decade, Ping An has built one of China’s broadest financial services and healthcare ecosystems, with more than 10 apps spanning insurance, banking, investment, healthcare, and senior care. But as the ecosystem expanded, navigating services across multiple apps became increasingly complicated.
Ping An is now moving from “apps as the customer interface” to “intent as the customer interface.” Its agentic AI solution, Ping An Express Service, brings more than 300 digital services across 40 high-frequency scenarios into a single conversational interface. By understanding what customers want to accomplish and helping them take action more quickly, the solution is simplifying customer journeys while lifting conversion and retention rates.
In this webcast, Eileen Tu, Product Lead for Ping An Express Service, and Dr. Ni, Healthcare Scientist at Ping An Technology, explore the strategy and technology behind the solution. Discover how Ping An is using agentic AI to turn customer intent into action—making financial and healthcare services faster, simpler, and more intuitive while delivering measurable business value.
Timecode
00:00-01:16
From App Maze to One Ping An: The strategic shift from more than nine separate apps across insurance, banking, healthcare, and senior care to a unified interface designed to reduce customer friction.
01:17-02:33
Intent as the New Interface: Replacing app-based navigation with one button and one dialogue box, allowing customers to express their needs in everyday language while AI coordinates services across the Ping An ecosystem.
02:34-05:15
Agentic AI & Cross-System Execution: How a single customer request triggers intent recognition, identity and permission verification, cross-system orchestration, compliance controls, and autonomous follow-through in seconds.
05:16-06:51
The Economics of Agentic AI: How Ping An Express Service increases customer engagement and higher-value conversions while reducing acquisition costs, with AI agents helping to generate more than RMB133 billion in sales in 2025.
06:52-09:05
The Data & Ecosystem Moat: Combining specialized finance and healthcare models with real-world scenarios and a data flywheel built on a decade’s worth of data to create a closed-loop capability that competitors cannot easily replicate.
09:06-11:50
Trust, Governance & Responsible Execution: The four safeguards that enable AI to take consequential actions—verified identities and permissions, controlled access to tools, risk-based human oversight, and complete audit trails.
11:51-13:13
From Reactive Service to Proactive Advice: The evolution of Express Service into an AI advisor that understands customers’ life stages, identifies protection gaps, and recommends portfolio adjustments before customers ask.
13:14-14:39
Full-Stack AI as a Growth Engine: How Ping An’s integrated AI infrastructure, proprietary databases, domain-specific models, and real-world applications create scalable competitive advantages—positioning AI as more than a chatbot or cost-efficiency tool.
FAQ
Q：What is Ping An’s AI Express Service?
A：Ping An’s AI Express Service is an agentic AI assistant that enables customers to initiate banking, insurance, healthcare and senior care services by expressing their needs through a single conversation.
Q：How does Ping An’s AI Express Service direct one customer request onto the right service pathway?
A：Customers express what they need in their everyday language, and the AI agents recognize the intent, verify identity and access permissions, coordinate across systems, check compliance and proactively advance tasks. Within seconds, the agents can process requests that once required several apps or phone calls.
Q：How is AI Express Service different from a chatbot or a general-purpose AI assistant?
A：A conventional chatbot primarily answers questions, but Ping An’s AI Express Service is designed to move authorized service requests toward completion. It can interpret customer intent, plan tasks and coordinate actions across systems and the service network. The capability is built on full-stack technology infrastructure, including Ping An’s domain-specific databases, specialized AI laboratories, and enterprise-scale computing and leading models developed for regulated financial and healthcare scenarios.
Q：What services and customer scenarios does AI Express Service support?
A：AI Express Service manages more than 300 digital services across 40 high-frequency scenarios, serving Ping An’s more than 250 million retail customers.
Q： What customer and business results have AI agents delivered?
A：Across Ping An’s broader operations, AI agents assisted RMB 133 billion in sales in 2025 by supporting demand analysis, personalized recommendations and sales communications. In Ping An’s health insurance business, AI-assisted cross-selling within the ecosystem has also generated higher average transaction values, while customer-acquisition costs were 30–45% lower than through external channels. In addition, Ping An P&C generated RMB 10.5 billion in claims savings through smart fraud detection in 2025, exceeding RMB 10 billion for the third consecutive year.
Q：Why can’t general-purpose AI models alone meet the needs of finance and healthcare?
A：General-purpose models can interpret language and generate content, but regulated financial and healthcare services require specialized knowledge, controlled access and integration with professional workflows. AI must understand areas such as underwriting, claims, credit risk, medical guidelines, disease management and senior care while meeting privacy, compliance and human-review requirements.
Ping An connects domain-specific models and data with real business systems, service teams, doctors and online-to-offline networks. This creates a closed loop. Applying AI in real-world scenarios generates feedback from customers and domain experts, which is used to continuously refine the AI model and improve service delivery.
Ping An connects domain-specific models and data with real business systems, service teams, doctors and online-to-offline networks. This creates a closed loop. Applying AI in real-world scenarios generates feedback from customers and domain experts, which is used to continuously refine the AI model and improve service delivery.
Q：How does Ping An keep its agentic AI safe, compliant and under human oversight?
A：Ping An applies four layers of governance – identity and access permission checks, strict system access controls, human oversight for high-risk tasks and comprehensive audit trails – so every AI action is authorized, controlled and fully traceable.