BeiDou ChatBI - Data Elements Driven Intelligent Multi-Scenario Decision Platform Qorus-NTT DATA Innovation in Insurance Awards 2026

Concept and objectives
Data-Driven Smart Operations is a long-term strategy of Ping An, aiming to enable 1/2 to 2/3 of employees to adopt data analysis as their main work method over the long term, thereby achieving data-driven operational decision-making. Currently, data analysis remains cumbersome and time-consuming, mean while, business decisions overly rely on expert experience, resulting in significant subjectivity—these are persistent industry pain points. To address these challenges, Ping An Life insurance has built a conversational ChatBI platform focused on empowering users to efficiently unlock data value. Features like intelligent data querying, root cause analysis, and smart decision-making are designed to tackle common struggles: difficulty accessing data, chaotic data usage, and slow analysis processes, this ultimately enabling everyone to become a data analyst. The self-developed conversational ChatBI platform reshapes operational management processes through a five-step methodology—“view data, define problems, identify causes, propose strategies, and trace changes”, this procedure seeks to transform traditional manual analysis processes into intelligent, end-to-end data interaction systems, enabling everyone to act as a data analyst.

Reasons behind
The project emerged from critical operational challenges in current practices. Traditional data analysis workflows are labor-intensive, slow, and often prone to delays, while decision-making continues to rely heavily on subjective expert judgment. Non-technical users face significant difficulties in interpreting complex datasets, and regulatory compliance requires efficient reporting mechanisms to meet inspection standards. These systemic inefficiencies—ranging from fragmented data accessibility to compliance risks—served as the driving force behind the development of an AI-driven solution aimed at democratizing data utilization across the organization.

State of competition
Our research revealed that competitors such as Huawei’s Pangu and Beiji Jiuzhang Suanshu were still in exploratory phases, lacking the industry-specific depth required for financial insurance applications. This gap prompted us to pursue self-development, positioning our solution as a first-of-its-kind innovation in financial and insurance technology. By focusing on domain-specific customization, we prioritized regulatory compliance, risk modeling, and user-centric workflows—setting our platform apart from generic AI solutions.

Sources of inspiration
The rise of ChatGPT inspired us to integrate AI into data governance, creating a hybrid system where users interact with data through natural language. By combining large language models (LLMs) with our data middleware, we enabled real-time analytics and automated root-cause analysis. For example, a user could ask, “Why did claims spike in Region X?” and instantly receive a breakdown through the metrics graph. This fusion of AI and governance not only accelerates decision-making but also establishes a new standard for innovation in the industry.

Departments involved
Ping An Life Architecture and Data Management Team, Ping An Technology AskBob Team, Ping An Life Testing Team.

Cross-functional collaboration involving Product, Architecture, Frontend & Backend, Data Development, Data Analysis, Testing, and other functions.

Main results so far
The project has been implemented in three phased milestones since early 2024. The first phase, completed between January and March 2024, focused on establishing conversational data-querying capabilities with core features such as intent recognition, multi-turn dialogue management, related question recommendations, and visual output generation. The second phase, executed from April to June 2024, introduced diagnostic analytical tools including anomaly detection, hierarchical problem decomposition, and attribution analysis. The ongoing third phase, spanning July to December 2024, is developing intelligent decision-making modules with automated strategy recommendations and regulatory compliance functionalities such as scenario-specific report generation for anti-fraud and consumer protection inspections.