70 results about "Enterprise level" patented technology

The invention relates to the technical field of natural language processing, in particular to an enterprise number asking system and method based on combination of a large language model and an NL2SQL, and the method comprises the following steps: filling a prompt project template with a natural language query request and a structured metadata context, and generating a standard prompt; the standard prompt is input into the large language model, deep semantic analysis is carried out, and an SQL query draft is generated; self-evaluation is carried out on the questions, and when the questions are recognized, clarified questions are generated and returned to the user; receiving feedback of a user, and iteratively optimizing the SQL query draft based on the feedback; performing grammar verification and security verification on the final SQL query draft to generate an executable SQL statement; sending the executable SQL statement to a target enterprise database for execution; and the information is visually displayed to a user. The problem that the SQL generated in the prior art deviates from the real intention of a user and cannot meet the requirements of enterprise-level applications for accuracy and reliability can be solved.

The invention provides an enterprise-level intelligent assistant-oriented multi-modal dialogue management and task execution system establishment method, which is characterized by comprising the following steps of Step 1, multi-modal data acquisition and preprocessing; step 2, carrying out cross-modal information alignment and fusion; 3, multi-mode dialogue management is carried out, and after data fusion is completed, the system enters a dialogue management stage; step 4, multi-modal task execution: in the dialogue management process, when a user initiates a specific task request, the system enters a task execution stage; automatically distributing and executing a plurality of tasks in the enterprise by using the fused multi-modal information; step 5, designing and optimizing a system architecture; and an integrated multi-mode intelligent assistant architecture is designed. By integrating multi-modal data processing and natural language understanding (NLU) technologies, the accuracy and efficiency of the enterprise intelligent assistant in task execution are remarkably improved. The system can process various modal information such as texts, voices and images at the same time, and seamless connection of cross-modal dialogue management and task execution is achieved.

The invention provides an enterprise-level unstructured knowledge governance-oriented method and a storage medium. The method comprises the steps that an initial document is input into a visual language model for layout analysis, and document structure information of a target document is obtained; segmenting the document structure information to generate text slices to be vectorized; metadata extraction is carried out based on the text slices, and structured knowledge content vectors are generated; and performing index configuration according to the knowledge content vectors to form enterprise-level knowledge governance standards. According to the method, the document structure is analyzed through the visual language model, the text slices are generated through the intelligent segmentation algorithm, the knowledge vector is constructed through metadata extraction, and the mixed index configuration is performed, so that the problems of inaccurate layout analysis, logic structure damage and low retrieval efficiency in the traditional technology are effectively solved, and the method has the advantage of improving the knowledge management precision and availability.

The invention discloses an enterprise-level agent system and a data processing method. The system mainly comprises a sensing module, a memory module, a planning module, a Routine agent planning engine, an arrangement engine and an execution module. According to the enterprise-level agent system provided by the embodiment of the invention, short-term memory, long-term memory and enterprise-level memory are combined through a complete closed-loop architecture of the sensing module, the memory module, the planning module and the execution module, and precipitation of personalized and organization-level knowledge is supported; and meanwhile, in combination with a standard business process template provided for an enterprise by a Routine agent planning engine and splitting of a task process by an orchestration engine, role agents are allocated for each sub-task, so that coordination and cross-system orchestration of complex tasks are realized, the construction threshold of intelligent applications is reduced, and the business efficiency is improved.

The invention provides an enterprise-level API gateway management method and system, and the method comprises the steps: API registration: an API registration platform receives input or batch-uploaded API metadata, and submits an auditing application after the registration is completed; aPI examination and approval: executing a differential examination and approval process according to the interface type, and exporting interface data after the examination and approval are passed; aPI release: importing the approved interface data into an API gateway management station, persisting the data to a relational database at the management station, and issuing a change notification to an API gateway; and updating the configuration of the API gateway: pulling and caching the database configuration data when the API gateway is started, continuously monitoring the data change of the management station, and updating the local cache in real time to take effect of a new routing strategy. According to the application, the reliability, the safety and the observability of the whole API management system can be improved.

The invention discloses an enterprise-level fraud behavior automatic identification platform and method based on artificial intelligence, and relates to the technical field of enterprise fraud data processing. The enterprise-level fraud behavior automatic identification platform and method based on artificial intelligence comprises the following steps: S1, obtaining enterprise fraud data, and preprocessing the enterprise fraud data; s2, constructing a role behavior graph, evaluating the risk connection closeness between behavior entities, updating the role behavior graph based on an evaluation result, and constructing a potential fraud edge set; s3, identifying a path to which each potential fraud edge belongs, analyzing causal contributions of behavior and fund variables to path risks, marking path causal field nodes and defining a path causal evidence chain; and S4, quantifying the interpretation intensity of each field in the path causal structure, identifying a dominant responsibility field, constructing a responsibility chain and outputting a structured attribution result. The problems that in the prior art, fraud recognition results cannot be explained, and evidence chains and responsibility traceability are lacked are solved.

The invention discloses an enterprise-level three-party dependency package security management and control system and method, and relates to the technical field of software supply chain security, and the enterprise-level three-party dependency package security management and control system comprises the following modules: a timed task scheduling module, an external network package pre-scanning module, a private service package monitoring module, a dependency graph construction module, a product management and control module and a notification display module. By establishing an external network packet pre-scanning mechanism, security scanning is performed and a blocking list is generated before a dependent packet enters an enterprise private server, and introduction of a packet containing high-risk vulnerabilities is blocked from the source; meanwhile, through the continuous monitoring of the private server package and the construction of the dependency relationship graph, the vulnerability discovery and the accurate positioning of the influence range of the stored dependency package are realized; and finally, performing hierarchical management and control on the affected products based on vulnerability levels, and realizing timely transmission and situation visualization of risk information through a notification display module. According to the invention, full-life-cycle safety protection from an external network source to internal products of an enterprise is realized, and the safety management level and risk response efficiency of the dependent package of the enterprise are effectively improved.

The invention discloses an enterprise-level recommendation method, system and device based on a large language model and a medium, and relates to the technical field of information recommendation. The method comprises the steps of obtaining a login ID and a query request of a user, and obtaining an enterprise identity of the user and historical behavior characteristics of the user based on the login ID; determining a user portrait of the user based on the historical behavior characteristics and the enterprise identity; determining a keyword corresponding to the user based on the user portrait; converting the keyword and the query request into a target vector; and searching in a preset vector database based on the target vector to obtain recommendation information, and converting the recommendation information into a visual recommendation page for display. By implementing the technical scheme provided by the invention, the deviation between the pushed information and the actual demand of an enterprise can be reduced, and the service effectiveness of enterprise-level information recommendation is improved.

The invention provides a chemical industry park risk grading early warning method and system. According to the method, an enterprise-level and park-level bidirectional coupling and dynamic and static combined chemical industry park risk grading early warning model is constructed. Wherein the enterprise-level risk model carries out dynamic quantitative scoring on an enterprise from five dimensions of inherent risk, operation state, management efficiency, emergency capability and government supervision, an enterprise output value correction coefficient is introduced, and evaluation deviation caused by scale difference is eliminated. And the park-level risk model realizes intelligent research and judgment on the overall risk of the park by weighting and fusing enterprise risks, public area monitoring data and special operation conditions. Accurate identification and graded early warning of chemical industry park risks are realized based on a double-layer early warning system of enterprise risk and park risk linkage, and the active safety guarantee capability is improved.

The invention relates to an enterprise-level SaaS operation control system based on role layering and authority matrix management. The enterprise-level SaaS operation control system is suitable for complex business scenes of multi-tenant and multi-role cooperation. The system comprises an identity recognition unit, a semantic meta-model subsystem, an authority matrix generation engine, a strategy gray release subsystem, a service instruction mapping module and a unified scheduling and auditing module. Constructing a dynamic permission matrix through a unified semantic token, and complementing permission configuration in combination with context information to form a policy version capable of gray deployment; authority operation is executed in a standard interface form after semantic mapping, log recording and audit control are carried out by a scheduling module, a strategy can be automatically frozen and rolled back when abnormal identification is carried out, and automatic evolution and safe closed-loop management of authority configuration are realized.

The invention discloses an enterprise-level MCP service calling method and device, equipment, a medium and a product, and relates to the field of enterprise-level artificial intelligence security services, and the method comprises the steps that an LLM interaction module generates an MCP tool calling list according to a user service request and a user token, and signs the token and the list; the MCP module verifies the signature; the authentication module obtains a field white list according to the token and the list; the MCP module signs the token, the list and the field white list; the service module verifies the signature and sends a result to the MCP module; the MCP module filters and desensitizes the result to obtain a security result, and signs the security result; the LLM interaction module verifies the signature, and inputs a security result and a user service request into a large language model to obtain a natural language reply, the application can realize the problems of dynamic minimum permission, trusted transmission and sensitive data compliance and desensitization, and meets enterprise-level permission control requirements.

The invention discloses an enterprise-level knowledge base information query system and method based on HMAC, and relates to the technical field of information security, the system comprises a client, an authentication module, a large language interaction module and a knowledge base module; the client is used for acquiring an account, a password and a retrieval request of a user; the authentication module is connected with the client and is used for receiving the account and the password and generating a user token; the large language interaction module is connected with the client and used for receiving the retrieval request, generating a query plan and signing the query plan and the user token to obtain a first signature packet; the knowledge base module is connected with the client and is used for receiving the first signature packet and signing again after verification is passed to obtain a second signature packet; and the big language interaction module is also used for receiving the second signature packet, and generating a query result through a second big language model in the big language interaction module after the verification is passed. According to the invention, the security of information transmission between the client and the knowledge base can be improved.

The present invention discloses an enterprise-level fraud automatic identification platform and method based on artificial intelligence, which relates to the technical field of enterprise fraud data processing. The enterprise-level fraud automatic identification platform and method based on artificial intelligence include the following steps: S1, obtaining enterprise fraud data and preprocessing the enterprise fraud data; S2, constructing a role behavior graph, evaluating the risk connection closeness between behavior entities, updating the role behavior graph based on the evaluation results and constructing a potential fraud edge set; S3, identifying the path to which each potential fraud edge belongs, analyzing the causal contribution of behavior and capital variables to path risk, marking the path causal field nodes and defining the path causal evidence chain; S4, quantifying the explanatory strength of each field in the path causal structure, identifying the dominant responsibility field, constructing a responsibility chain and outputting a structured attribution result. This solves the problem in the prior art that fraud identification results are unexplainable and lack evidence chain and responsibility tracing.

The invention discloses a method and a device for automatically skipping repeated approvers based on a global monitor, and aims to solve the problems of low efficiency, poor experience and the like caused by repeated approval in an enterprise-level approval process. According to the method, a repeated approver detection monitor is automatically injected into all user tasks in a process definition analysis stage by inheriting a BPMN analysis monitor of a process engine, and a process definition file does not need to be modified. When the task is created, the monitor reads the repeated skipping configuration defined by the process, judges whether the scene is a multi-person approval or countersigning scene or not, queries a historical task table to detect whether approvers are repeated or not, and automatically completes the task and sets a skipping mark variable if the approvers are repeated. The method has the advantages of zero invasion, flexible configuration, scene compatibility, traceability, high reliability and the like, and the approval efficiency and experience are effectively improved.

The application provides a component security vulnerability processing method and device, equipment and a storage medium. It relates to the technical field of application security. The method comprises the following steps: based on a full amount warehouse of component risk knowledge, marking the known risk enterprise-level components existing in the enterprise, the version range of the enterprise-level components, and the related applications having a dependency relationship with the enterprise-level components in the pre-constructed component panoramic view, the component panoramic view recording all components in the enterprise and the applications having a dependency relationship with each component; determining the risk level of the enterprise-level components based on the use of the enterprise-level components in each related application; determining the risk processing strategy for the enterprise-level components based on the risk level; and processing the enterprise-level components published in the enterprise component warehouse based on the risk processing strategy. The method of the application solves the problem of poor rectification effect when rectifying the enterprise-level components having security vulnerabilities.

A cross-enterprise identity management and authorization control method, device and medium belong to the technical field of permission management. The method comprises the following steps: S1, registering an enterprise space on a platform, and registering a user identity in the enterprise space; S2, defining the permissions of different application roles at the platform level, setting the specific post permissions at the enterprise level, associating the corresponding application roles based on the post of the user's belonging enterprise and allocating the corresponding access permissions, and generating a static permission baseline with multiple levels of constraints; S3, calculating the dynamic controlled range of user permissions based on the context attributes of the user and the space-time access strategy and identity access strategy of the platform resources, wherein the context attributes include time, space and user identity; and S4, combining the static permission baseline and the calculation result of the dynamic controlled range to calculate the real-time resource permission data of the user to perform real-time control of the resource permissions. The application provides a dynamic permission engine and a multi-level identity authentication system to reduce the risk of permission abuse.

The invention discloses an enterprise-level form generation method and system supporting low-code development, and relates to the technical field of enterprise form generation, and the method comprises the following steps: building a semantic security construction layer, analyzing an original character string logic expression generated by a low-code logic composer, analyzing the logic expression into a plurality of structured semantic nodes, and storing the structured semantic nodes in a database; the action range of the variables is limited based on the structured semantic nodes, so that the logic generation process is expanded in the closed action range; and constructing a dynamic action domain isolation container based on the generated structured semantic nodes, and distributing each structured semantic node to an independent context space. According to the method, closed operation of logic generation and execution is realized through semantic security construction and dynamic action range isolation, variable pollution is prevented, and semantic stability is ensured; through cooperation of namespace control and a limited execution environment, logic operation forms a safe closed loop, and the operation stability and safety of the low-code form are improved.

The invention relates to the technical field of artificial intelligence, and is used for solving the problems that an existing employee digital duplication system often neglects effective collection and analysis of task performance data of employees in a unit period and lacks a screening mechanism based on historical behaviors and task performance; the problems of high response delay, task omission and overtime completion of partial clones in task execution due to employees of a construction source are solved; the enterprise-level employee digital duplication system based on artificial intelligence comprises a server, and the server is further connected with a data acquisition module, a difficulty judgment module, a response analysis module, a reliability analysis module, a planning module and an effect verification module. Quantitative analysis is carried out through the process complexity, the employee response timeliness and the task reliability degree, and the intelligence and the accuracy of the digital duplication system are improved in combination with a judgment rule and an interval classification mechanism.

The invention discloses a group insurance underwriting method based on artificial intelligence. The method comprises the steps that S1, enterprise level and employee level multi-source data information submitted by an insurance buying enterprise is acquired and preprocessed; s2, constructing a composite risk feature set of an enterprise level and an employee level based on the data preprocessed in the S1; s3, performing correlation evaluation on the data preprocessed in the S1 based on the composite risk feature set constructed in the S2, screening out high-correlation features higher than a preset threshold value, and generating feature vectors according to the high-correlation features; s4, performing primary review on the feature vectors generated in the S3, and screening insurance policies which are obviously compliant or illegal; s5, constructing a multi-model fusion engine for risk prediction by adopting a Stacking integrated learning architecture; and S6, carrying out risk scoring on the insurance policy screened and compliant in the S4 by adopting the multi-model fusion engine constructed in the S5, and outputting a final underwriting conclusion. According to the method, multi-dimensional analysis is carried out on the underwriting data, and the risk assessment report is generated regularly, so that risk grading is carried out on insurance enterprises and customers.

The invention relates to the technical field of financial risk control, in particular to an enterprise-level risk base line management and control system and method based on a dynamic rule engine and multi-level workflow. The system is used for realizing real-time / batch dual-mode scanning of customer risk portraits, automatic hierarchical control of black and grey lists, full-process digital examination and approval of special case services and list transfer, and multi-dimensional visual analysis of risk data.

The embodiment of the application discloses a kind of clothing data processing method and equipment based on enterprise-level operating system, belong to clothing data processing technical field, solve clothing industry data entry link, due to the higher error rate of manual operation, affect the accuracy and completeness of industry data problem.The entity data recognition and format uniformity of clothing basic data are carried out by prepositioning multimodal pre-training model;Clothing basic data is input into different data modules, and the knowledge graph corresponding to entity data and different data modules respectively, to detect error data for clothing basic data;The repair value prediction of detected error data is carried out by data prediction filling model, and the clothing data to be fused is generated by repaired data;The feature vector corresponding to the clothing data to be fused and historical clothing data respectively is associated degree processing, and constructs clothing association relationship network;Based on clothing association relationship network, the clothing data to be fused is fused to the corresponding position of system.

An enterprise-level router comprises a shell (1), indicating lamps (2) are arranged on the front face of the shell (1), a heat dissipation opening (5) is formed in the side face of the shell (1), a liquid crystal panel is arranged on the periphery of the heat dissipation opening (5) and electrically connected with the indicating lamps (2) arranged on the front face of the shell (1), and a power interface (3), a USB interface (6), a plurality of Ethernet interfaces (4) and a management interface (9) are formed in the back face of the shell (1). A circuit board is arranged in the shell (1), a central processing unit (7) is arranged on the circuit board, and at least one LC oscillating circuit is arranged on a network cable between the Ethernet interface and the circuit board in the direction from the Ethernet interface to the circuit board. The use function is expanded, the working state is visually displayed, and the transmission rate of uploading and downloading can be further improved.

The invention provides an enterprise-level data integration processing method which is applied to a data integration processing system. The method comprises the steps of obtaining original business data of various heterogeneous data sources, and performing standardization processing to generate initial standardized data assets; carrying out ETL processing on the initial assets and storing the initial assets in a unified storage layer to generate intermediate standardized data assets; performing data management on the intermediate standardized data assets to generate target standardized data assets; unified scheduling and fault-tolerant control are carried out on execution tasks depended on by target assets through visual DAG arrangement, and continuous and stable generation of the execution tasks is guaranteed; and finally, performing data analysis and visual display based on the target standardized data assets, and generating an analysis result interface. The standardized and automatic management of the full life cycle of the data is realized, the problems of difficult integration of the multi-source heterogeneous data, low processing efficiency, uncontrollable quality, high operation and maintenance cost and the like are effectively solved, and the efficiency, stability, accuracy and collaboration of data processing are improved.

The application provides a data unloading method and device for enterprise-level data, an electronic device and a storage medium. The application relates to data processing technology, and is particularly suitable for an enterprise-level data unloading scenario. The method comprises the following steps: acquiring a source data group in a source data layer of a data warehouse, so as to determine an enterprise-level data dictionary followed by the source data group; then, acquiring an application data version of a business component in a business system; and according to the application data version and the enterprise-level data dictionary, unloading enterprise-level data to the source data group, so that the data types of the business component and the source data group are consistent. The application can solve the problems that the data unloading transition in the prior art depends on specific databases and the business component is not connected with the enterprise-level data dictionary.

The invention provides an enterprise-level cloud native PaaS management platform, and effectively solves the existing technical problems. The platform integrates a multi-cluster unified management module, and realizes cross-cluster resource dynamic optimization configuration and efficient collaboration by means of a federated cluster technology and an intelligent scheduling algorithm; according to the full-life-cycle application management system, tool chain and code detection are synchronously provided during development, automation of test deployment is achieved, and intelligent monitoring self-healing of operation and maintenance is achieved, so that service iteration is greatly accelerated. Meanwhile, the heterogeneous environment adaptation mechanism eliminates bottom layer difference through a resource abstraction layer and a self-adaptive plug-in, and seamless operation of the application is guaranteed. The security, stability and observability module constructs a multi-level protection, redundant architecture and a log, index and link comprehensive monitoring system. According to the platform, the resource utilization rate and the management efficiency are remarkably improved, the operation and maintenance cost is reduced, and a perfect solution is provided for safe, stable and efficient operation of enterprise-level applications in a cloud native environment.

The application discloses a kind of causal graph driven enterprise intelligent decision-making method and system, and the application belongs to enterprise management technical field.The method includes: by obtaining enterprise business process, operating index and unstructured text data, constructs enterprise-level causal graph;Based on the data to be decided data execution path reasoning, determine multi-hop causal path chain;Determine current correlation rule and decision suggestion in combination with expert rule base;Introduce the causal graph and rule label of different industries, based on structure similarity and semantic mapping rule, identify transferable causal path and correlation rule;Finally, the current and migration knowledge are fused, and enterprise decision execution scheme and causal reasoning result are generated.The scheme is based on the comprehensive application of current and migration knowledge, effectively enhances the scientificity, accuracy and interpretability of decision-making, while supporting dynamic updating and inter-industry knowledge sharing, significantly improves enterprise decision efficiency and response speed, helps enterprises to realize intelligent and fine management.

The invention relates to the technical field of artificial intelligence, in particular to an enterprise-level permission implicit risk identification method and system based on a large language model, and the method comprises the steps: collecting permission setting data from one or more target information systems; uniformly converting the permission setting data in different formats into a structured data set comprising preset core permission elements; inputting the structured data set into a large language model subjected to security domain knowledge fine adjustment, and executing conversion from codes to texts to generate permission description; executing a relationship extraction task by using a large language model subjected to security domain knowledge fine adjustment, extracting permission entities and relationships from the permission description, and constructing a global permission knowledge graph; and visualizing the global authority knowledge graph to generate an authority topological graph, traversing the knowledge graph in combination with a built-in risk rule base to identify hidden risks, and generating a semantic alarm. According to the invention, intelligent, transparent and visual enterprise-level authority management is realized, and various authority hidden dangers can be found in time.

This application provides an enterprise-level model validation method and an enterprise-level model validation device. The method involves acquiring user requirement information and the previous version of the data model, parsing the user requirement information to obtain the parsing results, generating the current data model based on the parsing results (including model documents and their implementation attributes), performing a difference analysis between the previous version and the current data model to generate difference analysis results (including structural and field changes), generating an incremental difference SQL script based on the difference analysis results, executing the incremental difference SQL script in the target database environment, and performing a consistency check between the table structure of the target database after execution and the current data model to obtain the validation results. This application solves the problems of low efficiency and low accuracy in existing model validation methods.

The invention belongs to the technical field of data storage, and discloses an enterprise-level data security storage system and method based on a block chain. Comprising a construction module used for determining multiple pieces of key data of an enterprise and constructing an enterprise block chain based on the multiple pieces of key data; the storage module is used for determining complete data corresponding to the multiple pieces of key data and storing the complete data in a preset database; the setting module is used for determining the demand of the project and establishing a data field based on the demand of the project; according to the invention, subsequent immature bodies can be conveniently combined with transmission requirements, a buffer period is provided for generation of new immature bodies, the requirements which do not conform to preset rules can be separated, malicious access or the requirements which do not conform to the preset rules can be avoided, data in the sub-data areas can be obtained through connecting the sub-data areas through the swimming ports, and the data transmission efficiency is improved. The data security of the off-chain data can be ensured when an enterprise establishes a project and needs to obtain the data.

The invention discloses an enterprise-level application operation and maintenance risk management and control method and device, a program and a storage medium. The method comprises the following steps: acquiring a plurality of risk characteristics of a target risk event; determining the risk type of the target risk event; determining a risk coping strategy for the target risk event according to the risk field and a preset risk coping strategy set; under the condition that the risk field is the known risk, obtaining operation environment data of the operation and maintenance carrier; determining a target diagnosis script according to the operation environment data, the risk coping strategy and a preset diagnosis script library; determining a target solution for the target risk event based on the target diagnosis script and a preset solution set; sending the target solution to the front end, and in response to a received confirmation instruction sent by the front end, generating a risk disposal script according to the target solution; and controlling the operation and maintenance carrier to operate the risk disposal script, and outputting a risk disposal result. According to the method, the risk blind area in enterprise-level application operation and maintenance can be reduced.