An AI-based process platform for the financial field

Abstract

The present invention provides an AI-based process platform for the financial field, which includes an architecture governance module, a business modeling module, and a technology modeling module; the architecture governance module is used to provide guidance for the business modeling module and the technology modeling module, the business modeling module is used to implement various types of business modeling in different business fields, and the technology modeling module is used to establish the way and format contract of API or service interaction between IT systems, and then design the specific processing logic for implementing the contract inside the IT system. The process platform of the present invention is a one-stop operation platform for business modeling in the requirements stage and technology modeling in the analysis and design stage, and can realize the two-way evolutionary generation of design and source code by docking with the development platform, improving the efficiency of asset management.

Description

Technical Field

[0001] The present invention relates to the technical field of software modeling, and in particular to a process platform for the financial field based on AI. Background Art

[0002] At present, the software R & D implementation processes applied in the field of fintech generally have complex process flows, emphasizing the derivation process, being theory - heavy, and having rigorous logic. They are suitable for the process modeling process of software requirements and analysis and design from scratch and gradually deriving. However, the process platform is based on new understandings and new implementation practices, and follows the principles of being practical and easy - to - use. It conducts comprehensive optimization and simplification for implementation difficulties and problems. It models and manages the digitalization of software requirements and analysis and design assets, and is a one - stop operation platform for business modeling in the requirements stage and technical modeling in the analysis and design stage. Moreover, it can achieve two - way evolutionary generation of design and source code by docking with the development platform, improving the efficiency of asset management.

[0003] Driven by digital transformation, in the face of the challenges of complex processes, the need for collaborative work, the development trend of intelligent manufacturing, the importance of knowledge management and inheritance, the goal of cost and resource optimization, the requirements of quality control and standardization, and the urgency of technological innovation and competition, enterprises urgently need a set of implementation processes to achieve digitalization, intelligentization, high - efficiency, and precision of process modeling, so as to improve design efficiency, optimize production processes, reduce costs and increase efficiency, and ultimately enhance market competitiveness.

[0004] The process platform undertakes the assets of the enterprise architecture and plays a guiding role in the development state. Traditional process platforms can generally only achieve one - way synchronization, that is, the architecture assets in the process platform can be unidirectionally synchronized to the implementation side, and code skeletons can be generated through self - developed IDE plugins. However, two - way synchronization cannot be achieved.

[0005] The implementation process is a complete set of methodologies for the entire software R & D life cycle that starts from business modeling (including process modeling, product modeling, entity modeling), takes the online switch as the end point of the first iteration, and can be continuously iteratively evolved, including processes, methods, templates, guidelines, specifications, etc. Following this process can ensure a global view and modelization of requirements, and can ensure that IT applications completely, correctly, and meticulously undertake business models. The process can cover all aspects of requirements, analysis, design, development, testing, deployment, and switching, as Figure 1 shown.

[0006] Currently, the software R & D implementation processes applied in the fintech field on the market generally have complex process flows, emphasizing the derivation process, being theory - heavy, and having rigorous logic. They are suitable for the process modeling of software requirements and analysis and design from scratch and gradually deriving. Moreover, traditional implementation processes have cumbersome procedures, high difficulty in getting started, and require a very good understanding of the relevant knowledge and methodologies of enterprise architecture to be proficiently used, and also have strict requirements for users' professional knowledge; when it comes to implementation, they often face the dilemmas of long cycles and large investments. In addition, the traditional implementation process is aimed at the modeling of enterprise - level full software systems, and it is extremely difficult to promote within the enterprise; the traditional implementation process focuses on the requirements stage of software R & D, and the automation degree of the connection from design to development often cannot meet the needs of implementation. Often, design is design and development is development, and design cannot well guide development, and the two will be decoupled. From the perspective of architecture governance, it is also impossible to well utilize the architecture assets in the process to manage the problems in the implementation process. Summary of the Invention

[0007] To solve the above problems, based on the basic principles of practicality and usability, the present invention invented a process asset platform in view of the implementation difficulties and problems, relying on rich practical experience and profound insights into industry pain points.

[0008] The present invention provides a process platform based on AI for the financial field, including an architecture governance module, a business modeling module, and a technology modeling module;

[0009] The architecture governance module is used to provide guidance for the business modeling module and the technology modeling module; the architecture governance module includes a business architecture unit and a technology architecture unit; the business architecture unit includes four model elements: product line, business domain, business capability, and business object; the technology architecture unit is used to define application components, and an application component is a set composed of a group of services with business internal connections and correlations;

[0010] The business modeling module is used to implement various types of business modeling in different business domains, including a process modeling unit, a product modeling unit, and an entity model unit;

[0011] The technology modeling module is used to establish the way and format contract of API or service interaction between IT systems, and then design the specific processing logic for implementing the contract inside the IT system.

[0012] Optionally, the process modeling unit includes three model elements: activity, task, and business rule;

[0013] An activity is triggered by an internal event or an external event and is a process to meet specific business purposes or respond to the demands of stakeholders;

[0014] A task is a container for a set of business rules that occur continuously and uninterruptedly in an activity with a clear business purpose;

[0015] Business rules are a business language that is clearly organized and semantically explicit from a business perspective and that uses natural language to structure and express business logic.

[0016] Optionally, the product modeling unit includes a product group, a base product, a product component, and a product condition;

[0017] Product groups are classified under product lines by product function or feature, usually grouping products by system application;

[0018] The basic product is an abstract product used to support innovation, which is clustered by a group of salable products with similar functions and business processing rules;

[0019] Saleable products are products that the enterprise independently sells and operates externally;

[0020] Product conditions are used to determine the business rules and restrictions of different products. They are a kind of restriction, constraint and limitation, and have perfect values ​​that can be used to describe product characteristics.

[0021] A product component is a product or service package provided by an enterprise to customers that cannot be sold separately and must be included in the saleable product for sale. It is composed of a set of product conditions.

[0022] The entity model involves business entities and their key attributes. Business entities are abstracted entities in the business to form independent and reusable units.

[0023] Optionally, the technical modeling module includes a contract determination unit and a processing logic unit;

[0024] The contract determination unit includes two model elements: interface and data entity; the types of interfaces include online interfaces that occur in real time and need to be processed immediately and require the system to respond in a short time, and batch interfaces that centrally process data according to specific trigger conditions within a specific time period.

[0025] Optionally, the interfaces and data entities of the technical modeling module have a certain connection relationship with the activities and business entities of the business modeling module, and the interfaces within the technical modeling module have a realization relationship with the online transaction and the batch transaction;

[0026] The process platform has a one-click generation function, which can generate corresponding interfaces based on activities, convert business entities into data entities, generate online transactions through online interfaces, generate batch transactions through batch interfaces, and automatically establish associations between assets.

[0027] Optionally, the process platform is equipped with a quick import mechanism for completing the process modeling and incorporation of a set of existing systems within a set event; the quick import mechanism includes business model import and technical model import;

[0028] The business model import takes the requirement documents of the existing systems as the data basis, constructs mapping relationships by deeply analyzing the document structure and the model elements of the business model, and realizes the efficient bottom-laying of the business model assets of the process platform by means of the general technology of importing offline Word documents into the Web server;

[0029] The technical model import directly targets the source code packages of the existing systems, analyzes the code structure and the model elements of the technical model to form mapping relationships, then polls and parses the Xml and Java files in the source code packages, and finally stores them accurately in the technical modeling module according to the mapping relationships.

[0030] Optionally, the process platform package has a user interaction page;

[0031] The user interaction page provides a workbench, process flow guidance, space configuration, and quick access;

[0032] The user interaction page also provides business architecture presetting, technical architecture presetting, and information standard entry, process model definition, product model definition, and entity model definition, contract design, implementation design, and design for docking with development;

[0033] The user interaction page also has the functions of statistical reports, requirement sheets, quick import of assets, permission management, and version management.

[0034] Optionally, iterative requirements for model assets are entered through the user interaction page, and combined with the model assets of the architecture governance module, business modeling module, and technical modeling module, a dedicated requirement sheet AIAgent is generated for the model assets of each version.

[0035] Optionally, there is a mapping relationship between the technical modeling module and the source code project files, and the source code of the system framework is automatically generated based on the design assets of the technical modeling module;

[0036] The design assets include online transactions, batch transactions, application service classes, basic service classes, data entities, data objects, custom SQL, error codes, and data dictionaries.

[0037] Optionally, the process platform is equipped with a consistency verification mechanism for verifying the consistency of the design assets and the source code project files, and when there is an inconsistency between the design assets and the system source code, the specific inconsistent information is located and displayed.

[0038] The AI-based process platform for the financial field of the present invention can combine with a large model for code interpretation to interpret and extract the function changes in the source code through an AI Agent based on the large model for code interpretation, and synchronize them backward to the architecture assets of the process platform. Thus, the problem that the architecture assets are out of sync with the code during the development of software engineering work is avoided.

[0039] The process platform accurately captures key assets from the processes and simplifies the derivation process, constructs clear and concise asset relationships, only retains the definitions of core key elements such as business fields, activities, tasks, and business entities that are necessary in the enterprise architecture process, removes a large number of intermediate process-style derivation processes, reduces the number of concept nouns in process modeling, thereby reducing the difficulty of getting started, and realizes the flexibility and scalability of asset management through the decoupled association of requirement assets and IT assets.

[0040] The process platform can efficiently manage the assets of mature target software products, convert traditional requirement documents and source codes into structured assets of business models and technical models, complete a set of process modeling assets within 30 minutes, greatly shorten the implementation cycle, and reduce resource investment.

[0041] The object of process modeling of the process platform is at the single domain level, rather than the entire software domain at the enterprise level in traditional implementation processes. It can first be piloted in a single system and then promoted to all systems of the enterprise to improve the acceptability of the enterprise.

[0042] The technical modeling of the process platform can be connected to the development platform. It can not only automatically generate development source codes through design assets, but also reverse generate design assets through source codes and with the help of a large model for code interpretation, realizing the two-way evolutionary generation of design assets and development source codes, ensuring the continuous optimization and upgrade of the process model, and also facilitating the implementation of process modeling.

[0043] The AI-based process platform for the financial field of the present invention is based on new cognition and new implementation practices, and comprehensively optimizes and simplifies according to the principle of being practical and easy to use, aiming at implementation difficulties and problems. It models and manages the digitalization of software requirements and analysis and design assets, is a one-stop operation platform for business modeling in the requirements stage and technical modeling in the analysis and design stage, and can realize the two-way evolutionary generation of design and source codes by docking with the development platform, improving the efficiency of asset management.

[0044] From the following detailed description of specific embodiments of the present invention in conjunction with the accompanying drawings, those skilled in the art will become more clear about the above and other objects, advantages and features of the present invention. Description of the Drawings

[0045] Upon reading the detailed description of the preferred embodiments below, various other advantages and benefits will become clear to those of ordinary skill in the art. The drawings are only for the purpose of showing the preferred embodiments and are not considered to be a limitation of the present invention. Moreover, throughout the drawings, the same reference numerals are used to represent the same components. In the drawings:

[0046] Figure 1 is a schematic diagram of the entire life cycle of software R & D;

[0047] Figure 2 is a schematic diagram of the process platform meta - model of the embodiment of the present invention;

[0048] Figure 3 is a schematic diagram of the rapid import process of the process platform business model of the embodiment of the present invention;

[0049] Figure 4 is a schematic diagram of the rapid import process of the process platform technology model of the embodiment of the present invention;

[0050] Figure 5 is a schematic diagram of the working process of the process platform requirement form AI Agent of the embodiment of the present invention;

[0051] Figure 6 is a partial effect diagram of the workbench of the embodiment of the present invention;

[0052] Figure 7 is a partial effect diagram of the architecture governance of the embodiment of the present invention;

[0053] Figure 8 is a partial effect diagram of the business modeling of the embodiment of the present invention;

[0054] Figure 9 is a partial effect diagram of the technical modeling of the embodiment of the present invention;

[0055] Figure 10 is a partial effect diagram of the consistency check of the embodiment of the present invention. Detailed Embodiments

[0056] The embodiments of the present invention will be described below with reference to the accompanying drawings. Those skilled in the art should understand that these embodiments are only for explaining the present invention and are not restrictive.

[0057] After a lot of industry research and example process deduction, the embodiment of the present invention provides an AI-based process platform for the financial field, including three modules: architecture governance module, business modeling module and technical modeling module. The architecture governance module is used to provide guidance for the business modeling module and technical modeling module. The business modeling module is used to implement various types of business modeling in different business fields. The technical modeling module is used to establish the method and format contract of API or service interaction between IT systems, and then design the specific processing logic of implementing the contract within the IT system, as shown in the following example. Figure 2 As shown. The AI-based process platform for the financial field in this embodiment clarifies the relationship between the technical modeling module and the development side. It not only defines the model elements and their attributes of each module, but also accurately describes the relationship between the model elements, providing a solid theoretical foundation and guiding framework for the construction and implementation of the process platform. Each module is described in detail below.

[0058] 1. Architecture Governance Module

[0059] The purpose of the architecture governance module is to provide clear guidance for business modeling and technical modeling. To this end, two sub-modules, business architecture unit and technical architecture unit, are carefully constructed in the architecture governance module.

[0060] The business architecture unit covers four model elements: product line, business domain, business capability and business object. In the business architecture unit, product line is the division of major product categories independent of organizational structure, customer segmentation or a specific channel, such as the bank's deposit product line and loan product line, which clearly define the bank's product categories. Business domain is the process of dividing and reflecting the overall value creation of the enterprise from the dual perspectives of customers and enterprises, such as the bank's personal deposit business domain and corporate deposit business domain, which accurately reflect the bank's service scope. Business capability is a set of capabilities that have clear business meaning and delivery value for the realization of business domains, such as inventory management and pricing management of banks, which together ensure the efficient operation of the business. Business objects are the structured embodiment of the core resources of the enterprise's business, such as the bank's personal current accounts and saleable products, which form the basis of bank operations.

[0061] The technical architecture unit focuses on the precise definition of application components. Application components are a set of services that have intrinsic connections and correlations with the business, such as the bank's deposit application components and public application components, which provide solid technical support for the enterprise's business. Through such an architectural design, architectural governance can promote the deep integration of business and technology in a more systematic and refined manner, laying a solid foundation for the sustainable development and innovation of the enterprise.

[0062] 2. Business Modeling Module

[0063] The business modeling module is further divided into three sub - modules: the process modeling unit, the product modeling unit, and the entity model unit.

[0064] The process modeling unit covers three model elements: activities, tasks, and business rules. An activity is triggered by an internal event or an external event and is a process to meet specific business purposes or address stakeholder demands. A task is a container for a set of business rules that occur continuously in time within an activity and have a clear business purpose. A business rule is a business language that clearly and semantically expresses business logic in a structured form using natural language from a business perspective.

[0065] The product modeling unit includes product groups, basic products, product components, and product conditions. A product group is classified by product function or feature under a product line and is usually grouped by system application. A basic product is an abstract product used to support innovation and is formed by clustering a group of salable products with similar functions and business processing rules. A salable product is a product that an enterprise sells and operates independently externally. A product condition is a restriction, constraint, and limitation used to determine the business rules and restrictions of different products and has a complete set of values that can be used to describe product characteristics. A product component is a product or service package provided by an enterprise to customers that cannot be sold separately and must be included in a salable product and consists of a set of product conditions.

[0066] The entity model unit involves business entities and their key attributes. A business entity is an abstraction of entities in business, forming an independent and reusable unit. Among them, the key attribute refers to the important entity attributes of a business entity, and here it can be changed to business entities and entity attributes.

[0067] The three sub - modules of the business modeling module achieve a close internal association through a carefully designed interaction mechanism. Specifically, the activities in the process modeling unit reference the basic products in the product modeling unit to ensure the accurate embedding of product information in the process; at the same time, the tasks in the process modeling unit reference the business entities in the entity model unit to ensure seamless docking of business logic and entity data. In addition, each activity is clearly defined under the business domain of the architecture governance module, and each task is accurately positioned within the business capabilities of the architecture governance module, thus realizing the precise guidance of the architecture governance module to the business modeling module.

[0068] III. Technical Modeling Module

[0069] The technology modeling module adopts the "contract first" design concept and conducts technology modeling from two parts: contract and implementation. That is, first establish the contract for the way and format of API or service interaction between IT systems, and then design the specific processing logic for implementing these contracts within the IT system on this basis. This design concept emphasizes the leading nature of the interaction protocol, ensures the standardization and consistency of inter-system communication, and provides a clear guiding framework for subsequent internal logic development.

[0070] That is to say, the technology modeling module includes a contract determination unit and a processing logic unit. Combining years of implementation experience in the industry, the contract determination unit designs two model elements: interface and data entity. The interface is further divided into an online interface that occurs in real time and requires immediate processing, demanding the system to respond within a short time, and a batch interface that processes data centrally according to specific trigger conditions within a specific time period. The data entity is the most basic unit for constructing data storage and is defined to handle the interaction method of directly reading database tables between IT systems. For example, the bank data warehouse system directly reads the background database table of the bank's core business system to extract data.

[0071] The processing logic unit of the technology modeling module involves four model elements: online transaction, application service class, basic service class, and batch transaction. The online transaction is an independent service for implementing the online interface in the contract part, including the definition of basic information such as transaction code and transaction name, input and output parameter definitions, and the call orchestration of application service classes. The application service class is a Java class that realizes a specific purpose and can be repeatedly called by online transactions and batch transactions. The basic service class is the smallest processing unit for implementing business rules and is also a Java class that can be repeatedly called by application service classes. The batch transaction is a batch processing service for implementing the batch interface in the contract part, including batch transaction code, batch transaction name, batch transaction type, description, input parameter definition, pre-batch job processing logic, normal processing logic, post-batch job processing logic, and exception handling logic. Moreover, the file batch transaction also includes the definition of file format.

[0072] The entire technology modeling module establishes an association relationship with the activities and business entities of business modeling through the model element of application component scope to achieve the integration of business and technology; and through non-essential weak associations, it successfully decouples from business modeling and realizes the scalability of the module. All model elements of technology modeling and elements of the development platform can be mapped one-to-one, laying a foundation for the docking between technology modeling and the development side. The application component scope refers to the scope of business modeling that a certain application component needs to implement.

[0073] The AI-based process platform for the financial field in the embodiment of the present invention also has a variety of functional mechanisms, which will be described in detail below.

[0074]

Rapid Import Mechanism for Process Platform

[0075] During the entire implementation cycle of process modeling, the largest investment lies mainly in the initial bedding work of modeling assets. This stage requires a large amount of manual adaptation and mechanical input. Therefore, the process platform of this embodiment is specifically designed with a set of rapid import mechanisms, aiming to complete the process modeling and management of a legacy system within 30 minutes, thereby significantly improving efficiency and reducing labor costs. In practical applications, a set of modeling assets will be formed after the implementation of process modeling. Assuming that model elements are compared to database table structures, then modeling assets are table data.

[0076] Optionally, the rapid import mechanism of the process platform based on AI for the financial field in this embodiment can be divided into two major links: business model import and technical model import.

[0077] The business model import takes the requirement documents of the legacy system as the data basis. By deeply analyzing the document structure and the model elements of the business model, mapping relationships are constructed, and with the general technology of importing offline Word documents into the Web server, the efficient bedding of the business model assets of the process platform is realized, as Figure 3 shown.

[0078] The technical model import directly targets the source code package of the legacy system, analyzes its code structure and the model elements of the technical model, forms mapping relationships, and then uses the parsing ability of the back-end of the process platform to poll and parse the Xml and Java files in the source code package, and finally accurately stores them in the technical modeling module according to the mapping relationships, thereby realizing the rapid and accurate import of the technical model. As Figure 4 shown.

[0079]

AI Agent for Process Platform Requirement Form

[0080] As the software system is successfully launched and switched, the model assets of its process modeling also enter a new iteration cycle. Through the function of the requirement form AI Agent of the process platform, it can not only effectively manage this iteration process, but also record the original requirements of each round in more detail and deeply analyze their potential impacts. The technical modeling part of the process platform can be understood as the general design and detailed design of the software system. The launch and switch refer to the operation and maintenance of the software system on the production environment and the switch to the iteration cycle. The model assets of process modeling are the model data formed after the completion of process modeling.

[0081] Users first pass through the Web page of the process platform (see Figure 6The platform then combines the model assets of the three modules of architecture governance, business modeling, and technical modeling to generate an exclusive requirement list AI Agent for each version of the model assets. After obtaining the original requirements, the AI ​​Agent can intelligently identify the purpose of the requirements, automatically analyze the key impact points on the model assets, and present multiple iteration plans for selection. Users can manually confirm and optimize the plans to achieve collaborative work with the requirement list AI Agent, and jointly promote the efficient and accurate iteration of the process platform model assets. Figure 5 shown.

[0082] For example, the deposit business model V1.0 and the loan business model V2.0 are two versions of model assets for different business modeling. Two AI Agents will be generated accordingly to identify the purpose of the entered iterative new requirements.

[0083] [Multi-space and multi-version design of process platform]

[0084] In order to ensure smooth development, testing, and production during the software system development process, it is usually necessary to prepare multiple operating environments to meet the needs of different stages. For the situation where multiple iterative requirements are distributed in different stages of development, the corresponding requirement version and source code version need to be released at each stage. In response to this development rule, the process platform has designed and implemented multi-space and multi-version functions.

[0085] In the process platform, the storage of model assets is clearly marked with the space and version number to which they belong. Space refers to the platform's tenant capabilities, one tenant per space, and the data between spaces is isolated from each other. The front-end query and display will also provide space and version number filtering options to ensure that the model asset data of each space viewed by the user is isolated from each other. Users can easily view the corresponding version of the model asset by switching between different versions, which greatly improves the flexibility and convenience of version management and asset viewing.

[0086] [Design of one-click generative conversion of process platform model elements]

[0087] There is a certain connection between the interface of the technical modeling module of the process platform, the data entity and the activities and business entities of the business modeling module, and at the same time, there is an implementation relationship between the interface inside the technical modeling module and the online transaction and batch transaction. Based on this design, the process platform of this embodiment realizes an efficient one-click generation function: the corresponding interface can be generated based on the activity, the business entity can be converted into a data entity with one click, the online interface can generate an online transaction with one click, and the batch interface can generate a batch transaction with one click, and the association relationship between these assets can be automatically established.

[0088] The specific conversion rules are as follows: basic information such as the name and description of an activity is accurately converted into the name and description of an interface, and its input and output are also synchronously mapped to the corresponding parameters of the interface; the name, description, and attributes of a business entity are carefully converted into the name, description, and fields of a data entity; at the same time, the interface code, name, description, input, and output of an online interface are skillfully converted into the transaction code, name, description, input, and output of an online transaction, realizing the smooth connection of the transaction process; the relevant information of a batch interface, including the interface code, name, description, input, and file template, is also accurately converted into the corresponding elements of a batch transaction, including the transaction code, name, description, input, and file format, thus ensuring the efficiency and standardization of batch transactions.

[0089] The system can automatically perform one-key generation and conversion of model elements with close relationships according to the set model element conversion rules, that is, corresponding interfaces can be generated with one key based on activities, business entities can be converted into data entities with one key, online interfaces can be generated into online transactions with one key, and batch interfaces can be generated into batch transactions with one key, and the association relationships between these assets can be automatically established. Then, the user can appropriately merge, split, and supplement the generated model elements, which quickly creates new model assets and greatly reduces the mechanical input workload.

[0090]

Process Platform Interface Design

[0091] As a tool for implementing processes, the process platform has a clear functional division based on each functional module in the process platform, which is reflected through the process platform interface design.

[0092] The process platform in this embodiment starts from three dimensions: user interaction experience, process service, and asset management, provides a user interface for realizing user interaction, and carefully designs the interfaces of each module, such as Figure 6 shown.

[0093] In terms of user interaction experience, the platform provides capabilities such as a workbench, process guidance, space configuration, and quick access, aiming to create an intuitive and convenient operation environment for users.

[0094] As the core of realizing the process meta-model, process service covers multiple important links. Among them, architecture governance includes business architecture presetting, technical architecture presetting, and information standard entry, as Figure 7 shown; business modeling involves process model definition, product model definition, and entity model definition, as Figure 8 shown; technical modeling includes contract design, implementation design, and design for docking with development, ensuring the comprehensiveness and accuracy of process service, as Figure 9 shown.

[0095] Asset management has functions such as statistical reports, requirement forms, quick asset import, permission management, and version management, providing strong support for the effective management and utilization of assets.

[0096]

Source Code of the Design Asset Automatic Generation System Framework Based on Process Platform Technology Modeling

[0097] The model elements of the technology modeling module can be mapped one-to-one with the elements of the development platform (the downstream of the process platform, specifically referring to the IDE integrated with plugins that can dock with the process platform). The model elements of the technology modeling module can be mapped one-to-one with the elements of the development platform. Therefore, the source code of the design asset automatic generation system framework can be generated based on the model assets of the technology modeling module of the process platform technology modeling module.

[0098] The design assets formed by the model elements of the technology modeling module mainly include online transactions, batch transactions, application service classes, basic service classes, and data entities; the configuration assets mainly include data objects, custom SQL, and error codes. Combining these two parts and the data dictionary required by the design assets, the system jointly generates the corresponding system framework source code. The specific file mapping relationship is shown in Table 1 below, the Design Asset and System Source Code File Mapping Relationship Table.

[0099] Table 1 Design Asset and System Source Code File Mapping Relationship Table

[0100]

[0101]

[0102]

The Design Assets of Process Platform Technology Modeling are Reverse Generated by the System Source Code with the Aid of AI Capabilities

[0103] Based on the mapping relationship between the design assets of the technology modeling module and the system framework source code, the process platform can access the system engineering source code on gitlab and parse it in turn according to a total of ten categories: online transactions, ordinary batch transactions, file batch transactions, application service classes, basic service classes, data entities, data objects, custom SQL, error codes, and data dictionaries. During the parsing process, the platform deeply explores the file structure and service definitions, and with the help of the AI Agent intelligent code interpretation logic based on the large model of code interpretation, converts them into natural language and refines them, and finally constructs the complete technology modeling design assets. Among them, the system framework source code refers to the framework of the source code engineering file. For example, there are only java class files, names, and input and output declarations, without the specific implementation business logic code of the class, while the source code engineering file is the source code that can only run and execute the business logic.

[0104] Specifically, the platform parses the [transaction code.flowtrans.xml] file to extract the name, description, input and output of the online transaction and the orchestration call to the application service class; interprets the [transaction code.java] file through the AI ​​Agent to describe the coding logic of the orchestration call method in natural language. Similarly, the [transaction code.batch_tran.xml] file is parsed to generate the batch transaction code, batch transaction name, batch transaction type, description, and input parameter definition of ordinary batch transactions, and the code interpretation AI Agent reads the [batch transaction code DataProcessor.java] to generate the natural language expression of the batch job pre-processing logic, normal processing logic, batch job post-processing logic, and exception processing logic. Similarly, the [batch transaction code.file_batch_tran.xml] file is parsed to generate the batch transaction code, batch transaction name, batch transaction type, description, input parameter definition, and file format definition of file batch transactions, and the code interpretation AI Agent reads the [batch transaction code ReadFileProcessor.java] and [batch transaction code WriteFileProcessor.java] files to generate the natural language expression of the file batch pre-job processing logic, normal processing logic, batch job post-processing logic, and exception processing logic. In addition, the platform uses the interpretation AI Agent to parse the XML and Java files related to the application service class and the basic service class, and generates the class code, name, implementation class name, description, method list and natural language processing logic of the application service class and the basic service. At the data level, the platform parses the [table name.tables.xml] and [data object code.c_schema.xml] files to generate the code, name, attribute list and other information of the data entity and data object respectively. At the same time, it parses the [custom SQL code.nsql.xml], [error code code.error.xml] and data dictionary related XML files to comprehensively generate detailed asset information of custom SQL, error code and data dictionary.

[0105]

Consistency check mechanism for process platform technology modeling assets and system source code

[0106] The system source code development of the development platform strictly follows the design assets formed by the technical modeling of the process platform. To ensure the rigor of this undertaking process, the embodiment of the present invention provides a set of precise docking mechanisms, namely, consistency verification mechanisms. The consistency verification mechanism can quickly locate the specific inconsistency points when there is inconsistency between the technical modeling assets and the system source code. Then, through a highly interactive display interface (such as a user interaction page), a clear and intuitive view is provided to the user, enabling them to easily and accurately identify and locate inconsistencies, thereby ensuring a smooth and efficient development process.

[0107] The current design assets include ten types of assets, such as online transactions, ordinary batch transactions, file batch transactions, application services, basic services, data entities, data objects, custom SQL, error codes, and data dictionaries. The consistency check interface conducts a detailed check and comparison of the design assets and system source code according to these ten types, and highlights the differences. During the consistency check process, if the system source code exists but the design asset is missing, the corresponding asset will be marked as "to be added", indicating that it needs to be supplemented; conversely, if the design asset exists but the system source code is missing, it will be marked as "to be deleted", indicating that it needs to be removed. For the situation where both exist but the detailed information such as encoding and name is inconsistent, it will be marked as "to be modified", indicating that it needs to be updated. Users can quickly filter according to the asset type and difference type (to be added, to be deleted, and to be modified) through the multi-dimensional filtering function of the consistency check interface, which is convenient for concentrating on viewing the differences of a certain category, so as to efficiently manage assets.

[0108] In addition, the consistency check system supports directly uploading the source code package for parsing, and can also directly access the source code on GitLab for reading, providing a flexible way to access the source code. At the same time, using the ability of scheduled tasks, the triggering method of consistency comparison has realized three methods: manual immediate trigger, specified time point trigger, and periodic trigger at time intervals, to meet the check requirements in different scenarios and further improve the intelligence and automation level of asset management. As Figure 10 shown.

[0109]

Quasi-consistency mechanism for process platform technology modeling assets and system source code

[0110] Relying on the consistency check interface, the preservation mechanism of the process platform can quickly ensure a high degree of consistency between the technical modeling assets and the system source code. For the design assets marked as "to be added", "to be deleted", and "to be modified" on the interface, the platform intelligently reads the system source code, accurately obtains the relevant data, and automatically executes the "add", "delete", and "modify" operations of the design assets to ensure that the technical modeling assets and the system source code are always consistent. In addition, the platform also supports users to batch select different files, and through the one-key synchronization function, quickly complete the consistency update of multiple files, further improving the convenience and accuracy of asset management.

[0111] The above is only the preferred implementation manner of the present invention. The protection scope of the present invention is not limited to the above embodiments. All technical solutions within the idea of the present invention belong to the protection scope of the present invention. It should be pointed out that for those of ordinary skill in the art in this technical field, several improvements and refinements made without departing from the principle of the present invention should also be regarded as the protection scope of the present invention.

Claims

1. An AI-based process platform for the financial field, characterized in that Includes architecture governance module, business modeling module and technical modeling module; The architecture governance module is used to provide guidance for the business modeling module and the technical modeling module; the architecture governance module includes a business architecture unit and a technical architecture unit; the business architecture unit includes four model elements: product line, business domain, business capability and business object; the technical architecture unit is used to define application components, which are a set of services with inherent business connections and correlations; The business modeling module is used to implement various types of business modeling in different business fields, including process modeling unit, product modeling unit and entity model unit; The technical modeling module is used to establish the interaction mode and format contract of APIs or services between IT systems, and then design the specific processing logic within the IT system to implement the contract; The process platform is provided with a rapid import mechanism for completing the process modeling and management of a set of stock systems within a set event; the rapid import mechanism includes business model import and technical model import; Business model import uses the demand documents of the stock system as the data basis, builds a mapping relationship through in-depth analysis of the document structure and the model elements of the business model, and uses the general technology of importing offline Word documents into the Web server to achieve efficient foundation of the business model assets of the process platform; The technical model import directly targets the source code package of the stock system, analyzes the code structure and model elements of the technical model, forms a mapping relationship, and then polls and parses the Xml and Java files in the source code package, and finally accurately stores them in the technical modeling module according to the mapping relationship; The technical modeling module includes a contract determination unit and a processing logic unit; the contract determination unit includes two model elements, namely, an interface and a data entity; the types of interfaces include online interfaces that occur in real time, need to be processed immediately, and require the system to respond in a short time, and batch interfaces that centrally process data according to specific trigger conditions within a specific time period; the interfaces and data entities of the technical modeling module have a connection relationship with the activities and business entities of the business modeling module, and at the same time, the interfaces within the technical modeling module have an implementation relationship with online transactions and batch transactions; the process platform has a one-click generation function, which generates corresponding interfaces based on activities, converts business entities into data entities with one click, generates online transactions with one click from online interfaces, generates batch transactions with one click from batch interfaces, and can automatically establish associations between assets; wherein, activities are model elements in the process modeling unit, which are triggered by internal or external events and are processes that meet specific business purposes or respond to stakeholder demands; The technical modeling module and the source code engineering file have a mapping relationship, and the source code of the system framework is automatically generated based on the design assets of the technical modeling module; the design assets include online transactions, batch transactions, application service classes, basic service classes, data entities, data objects, custom SQL, error codes, and data dictionaries.

2. The process platform according to claim 1, characterized in that, The process modeling unit includes three model elements: activities, tasks, and business rules; Activities are triggered by internal or external events to meet specific business objectives or respond to stakeholder demands. A task is a container for a set of business rules that occur continuously in time during a certain activity and have a clear business purpose; A business rule is a business language that is well - structured and semantically clear from a business perspective and uses natural language to structure the expression of business logic.

3. The process platform according to any one of claims 1-2, characterized in that The process platform package has a user interaction page; The user interaction page provides a workbench, a process flow guide, space configuration, and quick access; The user interaction page also provides business architecture preset, technical architecture preset, and information standard entry, process model definition, product model definition, and entity model definition, contract design, implementation design, and design for docking with development; The user interaction page also has the functions of statistical reports, requirement sheets, quick asset import, permission management, and version management.

4. The process platform according to claim 3, characterized in that, Through the user interaction page, the iterative requirements of model assets are entered, and combined with the model assets of the architecture governance module, business modeling module, and technical modeling module, a dedicated requirement sheet AIAgent is generated for the model assets of each version.

5. The process platform according to claim 1, characterized in that, The process platform is provided with a consistency verification mechanism for verifying the consistency of design assets and source code engineering files. When there is an inconsistency between the design assets and the system source code, the specific inconsistent information is located and displayed.

6. The process platform according to claim 1, characterized in that, The product modeling unit includes product groups, basic products, product components, and product conditions; Product groups are classified by product function or characteristics under a product line; A basic product is an abstract product used to support innovation, which is formed by clustering a group of salable products with similar functions and business processing rules; A salable product is a product that an enterprise sells and operates independently outside; Product conditions are used to determine the business rules and restrictions of different products, which are a kind of restrictions, constraints, and limitations with complete value ranges for describing product characteristics; A product component is a product or service package provided by an enterprise to customers that cannot be sold separately and must be included in a salable product and consists of a group of product conditions.

Images