Innovations and Applications of Cognitive Decision-Making Intelligence

Introduction

The project was developed by Ant Group over eight years of technological research, which solved many technical problems faced by artificial intelligence in the implementation of life services, finance, healthcare and other industries. Significant breakthroughs were made in key technologies such as Large-scale Models, Graph Learning, Knowledge Graph, Operational Optimization, and Al agents, resulting in significant industrial application effects.

The Project Has Overcome the Challenges of Applying Al in Industries such as Life Services and Healthcare

The project has overcome three major challenges in the Applications of artificial intelligence in industries such as life services, finance, and healthcare: weak domain cognitive ability, poor complex reasoning ability, and difficult end-to-end landing. It has achieved multiple technological innovations:

1.We developed a trillion-parameter trustworthy large model foundation, which in conjunction with hundreds of billions of specialized corpus and tens of thousands of industry tools, has created large models for fields such as healthcare and finance. This has significantly enhanced the cognitive and operational capabilities of large models in vertical industries.

2. We developed a high-performance Knowledge Decision Engine. Through innovations in techniques such as graph learning, knowledge graphs, and operational optimization, real-time inference and decision-making have been realized based on trillions of user behaviors, industry knowledge, and expert experience.

3. We developed an industrial Al agent application platform, supporting the rapid construction of specialized Al agents, and bridging the last mile of Al application.

The platform supports several applica-tions, including Al Life Assistant "Zhi Xiao-bao", Al Financial Manager "Ma Xiaocai", and the first digital hospital companion "Angel".

The project has been granted over 100 authorized invention patents, published more than 50 top-tier papers, and some achievements have won multiple authoritative awards such as the First Prize of Zhejiang Science and Technology Progress Award, the First Prize of ClE Science andTechnology Progress Award.

The Project Has Been Widely Applied in Industries such as Medical, Finance

As of now, the project has been widely applied in more than 1000 institutions in industries such as life service, healthcare, and finance. As follows:

1. In the field of life services, the project has developed the first service-oriented Al native application "Zhi Xiaobao"

2. In the healthcare fields, the project has served multiple hospitals and institutions such as Shanghai Renji Hospital, Shanghai General Hospital, and so on. The digital hospital companion "Angle" has provided services to over 1000 healthcare institutions in Zhejiang, helped reduce the consultation volume for manual guidance by 50%.

3. In the financial decision-making field, the project has significantly improved the ability to depict key portraits and credit risks of micro-enterprises, achieving fine intelligent de-cision-making for access, loan limits, and interest rates, effectively alleviating the financing difficulties of micro-enterprises.

The Project Has Played a Positive Role in Promoting the Opening Up and Coordinated Development of the Industry

By promoting standards development, open industry datasets, and other means, the project has played an active role in promoting the opening and collaborative development of industries, supporting industry digital transformation.

1. Hosting and participating in more than ten international/national standards, covering knowledge graphs, graph learning, large models, and other technological fields.

2. Build or collaborate with industry partners to jointly build multiple datasets and open them to the entire industry, helping

Al empower the industry.