A process node-based orchestration tool

Through the orchestration tool based on process nodes, combined with technologies such as intelligent identification, adaptive execution and multi-dimensional data analysis, the problems of single functions of traditional tools and insufficient data processing capabilities are solved, flexible optimization and security management of processes are realized, and execution is improved. Efficiency and scientific decision-making.

CN118672646BActive Publication Date: 2024-12-03ZHEJIANG ZHENGYUAN ZHIHUI TECH CO LTD
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
CN202411172946.6
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-08-26
Publication Date
2024-12-03
Estimated Expiration
2044-08-26

AI Technical Summary

Technical Problem

Traditional process orchestration tools have single functions, which are difficult to meet complex business needs, have low execution efficiency, weak data processing capabilities, and are unable to deeply explore process relationships and causal connections, resulting in a lack of scientific basis for optimization.

Method used

Provides a process node-based orchestration tool, including intelligent identification and dynamic building modules, adaptive execution engine, real-time feedback and dynamic adjustment mechanisms, multi-dimensional data analysis, cross-platform integration, permission management and security audit, artificial intelligence assistant and automation The test module combines edge computing, machine learning and blockchain technology to achieve flexible process construction, real-time optimization and security management.

Benefits of technology

Significantly improve work efficiency and scientific decision-making, reduce risks and errors, enhance user experience and system stability, adapt to diverse needs and environments, and provide profound insights and strong decision-making basis.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118672646B_ABST
    Figure CN118672646B_ABST
Patent Text Reader

Abstract

The present invention discloses a process node-based orchestration tool, which relates to the technical field of service orchestration, and includes the following modules: an intelligent identification and dynamic construction module: including: an intelligent identification unit; a dynamic construction unit; an adaptive execution engine module: introducing edge computing capabilities, directly processing node tasks that require rapid response on edge devices, reducing network delays, and improving execution efficiency; a real-time feedback and dynamic adjustment mechanism module: including a real-time feedback unit; a dynamic adjustment mechanism unit: introducing an A / B testing mechanism, and testing different solutions in parallel during the dynamic adjustment process; a multi-dimensional data analysis module; a cross-platform and system integration capability module; In the present invention, the orchestration tool is rich in functions and the modules are integrated and coordinated, which significantly improves the efficiency of process construction and execution, has powerful data processing and analysis capabilities, deeply explores process relationships, provides a scientific basis for decision-making, and adapts to diverse needs and environments.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of service orchestration, and in particular to an orchestration tool based on process nodes. Background Art

[0002] In today's era of rapid digitalization and informatization, the complexity and diversity of various business processes are increasing. Traditional process orchestration tools often have many limitations and are unable to meet the ever-changing business needs.

[0003] Previous process orchestration tools usually have relatively simple functions, lack of effective integration and coordination between modules, resulting in low efficiency in process construction and execution. When faced with complex business scenarios, they cannot flexibly adjust and optimize dynamically according to business changes, resource availability and user-specific requirements. At the same time, traditional tools are weak in data processing and analysis, making it difficult to deeply explore the complex relationships and causal relationships in the process, and unable to provide strong support for decision-making, resulting in a lack of scientific basis for process optimization.

[0004] In view of this, this application is hereby filed. Summary of the invention

[0005] The purpose of the present invention is to provide a process node-based orchestration tool to solve the problems raised in the above background technology.

[0006] To solve the above technical problems, the present invention provides a process node-based orchestration tool, including the following modules: an intelligent identification and dynamic construction module: including: an intelligent identification unit: extracting key features and elements from the input process information; a dynamic construction unit: capable of generating a flexible and optimized process architecture in real time, dynamically adjusting and constructing process nodes based on the identified process information and specific rules and strategies; an adaptive execution engine module: introducing edge computing capabilities, directly processing node tasks that require rapid response on edge devices, reducing network delays, and improving execution efficiency; a real-time feedback and dynamic adjustment mechanism module: including a real-time feedback unit; a dynamic adjustment mechanism unit: introducing an A / B testing mechanism, testing different solutions in parallel during the dynamic adjustment process, and selecting the best solution driven by data; a multi-dimensional data analysis module; a cross-platform and system integration capability module: providing a low-code / no-code integration tool, supporting non-technical personnel to quickly configure system integration through a graphical interface; and supporting cloud-native deployment at the same time, ensuring that the orchestration tool can flexibly adapt to different cloud environments and improve scalability and flexibility; a permission management and security audit module; an artificial intelligence assistant integration module; an automated testing and verification framework module: introducing chaos engineering experiments, simulating extreme conditions to test system resilience.

[0007] Furthermore, the real-time feedback unit includes intelligent early warning, which uses machine learning algorithms to predict potential risks and issue early warning signals in advance, and gives preventive adjustment suggestions.

[0008] Furthermore, the multi-dimensional data analysis module: expands the functionality of the data-driven decision optimization system, introduces network analysis tools to reveal the complex relationships between process nodes, and provides comprehensive and in-depth insights and decision-making basis for process optimization.

[0009] Furthermore, the permission management and security audit module includes: a permission management unit based on blockchain technology, which records tamper-proof logs of key operations and data changes to improve audit transparency and credibility; and a dynamic access control unit, which dynamically adjusts access rights according to user behavior and environmental changes.

[0010] Furthermore, the artificial intelligence assistant integration module includes: an intelligent chat robot unit or a virtual assistant unit with adaptive learning capabilities; and an emotional interaction unit.

[0011] Furthermore, the automated testing and verification framework module: builds an automated testing and verification framework unit that includes the CI / CD process to ensure that each code change can automatically trigger testing and verification, reducing human errors and online risks.

[0012] Furthermore, it also includes a comprehensive monitoring and performance optimization module: including a comprehensive monitoring unit: real-time monitoring of the entire process of process execution, used for performance indicators, resource utilization and user feedback, etc., automatically triggering performance optimization measures to ensure efficient and stable operation of the system.

[0013] Furthermore, it also includes user-defined extension and plug-in support modules: it provides user-defined extension interfaces and plug-in support, allowing users to develop or integrate new functional modules according to business needs, thereby enhancing the flexibility and scalability of the orchestration tool.

[0014] Compared with the prior art, the present invention has the following beneficial effects:

[0015] Significantly improve work efficiency and execution results: Through the synergy of the adaptive execution engine and intelligent recognition and dynamic construction modules, the process and execution logic can be optimized according to the actual situation, effectively avoiding unnecessary steps and waiting time, greatly enhancing work efficiency and quality, and making business processing faster and more accurate.

[0016] Effectively reduce risks and errors: The real-time feedback and dynamic adjustment mechanism can provide early warning of potential risks, and the automated testing and verification framework can reduce human errors. In addition, the permission management and security audit modules can ensure the security of data and operations, which can comprehensively reduce the risks in business operations and the possibility of errors.

[0017] Deeply optimize decision-making and processes: The multi-dimensional data analysis module can deeply explore the complex relationships and causal connections in the process, providing decision-makers with a deep and comprehensive insight perspective and strong basis, which effectively promotes the optimization and improvement of the process, making decisions more scientific and reasonable, and the process more efficient and smooth.

[0018] Greatly enhanced user experience: The operation guidance and emotional interaction functions provided by the artificial intelligence assistant integration module greatly improve the user's satisfaction and convenience during use, allowing users to experience more intimate and humane services.

[0019] Excellent adaptability to diverse needs and environments: The cross-platform and system integration capability modules enable the tool to easily adapt to a variety of different systems and cloud environments, while the user-defined extension and plug-in support modules accurately meet personalized business needs, demonstrating wide applicability and being able to play a role in a variety of complex scenarios.

[0020] Solid guarantee of system stability and reliability: The comprehensive monitoring and performance optimization module can monitor the system's operating status in real time, automatically trigger optimization measures, and ensure that the system always maintains an efficient and stable operating state, providing reliable technical support for the continuous development of business. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 The figure is a principle block diagram of a process node-based orchestration tool. DETAILED DESCRIPTION

[0022] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention. Embodiment 1

[0023] A process node-based orchestration tool used in the generation, storage, sharing and analysis of electronic medical records.

[0024] See also Figure 1 , the present invention provides a technical solution:

[0025] See also Figure 1As shown, a process node-based orchestration tool includes intelligent recognition and dynamic construction modules: including: intelligent recognition unit: through algorithms and models, it deeply analyzes the input process information and extracts key features and elements; dynamic construction unit: it can generate a flexible and optimized process architecture in real time based on the identified process information and specific rules and strategies. It can dynamically adjust and construct process nodes according to business changes, resource availability and user-specific requirements; this module can automatically identify the medical record template requirements of different departments and different diseases.

[0026] The following are examples:

[0027] Data collection and annotation: First, collect a large number of medical record samples from different departments (such as cardiology, surgery, etc.) and different diseases, and annotate these samples in detail. For example, for cardiology medical records, annotate key information such as common symptom descriptions, diagnostic items, and treatment plans; for surgical medical records, annotate important content such as the type of surgery, preoperative preparation, and postoperative care.

[0028] Feature extraction and model training: Use natural language processing technology to extract key features from medical records. For example, the medical record text is converted into a computable vector form through methods such as bag-of-words model and TF-IDF. Then, use machine learning algorithms (such as decision trees, random forests, neural networks, etc.) to train the model. For example, taking neural networks as an example, a multi-layer neural network structure can be constructed, and the labeled medical record feature vectors can be input. The network parameters are continuously adjusted through the back propagation algorithm to improve the model's recognition accuracy for medical records of different departments and diseases. Neural network model training example: The input layer receives the feature vector of the medical record, for example, containing 50 features. Multiple hidden layers can be set in the middle, such as 2 hidden layers, containing 100 and 50 neurons respectively. The output layer is set according to the prediction task. For example, the prediction department can be set to 5 neurons, corresponding to 5 different departments. During the training process, the labeled medical record feature vectors are input into the network, and the connection weights between neurons are adjusted through the back propagation algorithm so that the output of the network gradually approaches the actual annotation results. For example, among a batch of medical records, after the medical records that are truly labeled as internal medicine are input into the network, the network's initial output may be surgery, but by continuously adjusting the weights, the final output gradually approaches internal medicine, thereby improving the accuracy.

[0029] Establish a department and disease classification system: formulate clear rules for the classification of departments and diseases. For example, medical records of cardiology are classified into one category, which is further subdivided according to common cardiovascular diseases (such as coronary heart disease, arrhythmia, etc.); surgical records are subdivided according to the type of surgery (such as general surgery, orthopedic surgery, etc.).

[0030] Dynamically construct medical record framework and required items: When a new medical record is input, the model selects the corresponding framework from the pre-set template library based on the department and disease information identified by the implementation example. For example, for the medical record of coronary heart disease in the cardiology department, a framework containing required items such as electrocardiogram results, myocardial enzyme indicators, and drug treatment records is dynamically constructed; for the medical record of fracture surgery in the surgery department, a framework containing required items such as operation time, fracture site description, and fixation method is constructed.

[0031] Real-time update and optimization: As medical practice develops and new medical record data accumulates, models and templates are continuously updated. For example, if new treatment methods or diagnostic indicators emerge, they are promptly incorporated into the medical record templates of the corresponding departments and diseases to ensure the accuracy and practicality of the templates.

[0032] Collaborate with clinical experts for verification: Regularly invite clinical experts to evaluate and verify the automatically constructed medical record framework and required items. For example, organize an expert meeting every month to jointly review the results generated by the model and adjust and improve the model based on the feedback from experts.

[0033] The intelligent recognition and dynamic construction module uses a series of advanced technologies and methods, such as natural language processing technology to understand process descriptions, machine learning algorithms to predict and optimize construction strategies, and distributed computing frameworks to improve processing speed and efficiency. Through continuous learning and improvement, the intelligent recognition and dynamic construction module can continuously improve its accuracy and flexibility, providing strong support for the intelligence, humanization and security of the entire process orchestration tool. Through the comprehensive application of the above methods and continuous optimization and improvement, it can gradually realize the intelligent recognition of medical record template requirements for different departments and different diseases, and dynamically build accurate and practical medical record frameworks and required items.

[0034] Adaptive execution engine module: The adaptive execution engine module integrates a self-learning algorithm, which can continuously collect feedback during the execution process, automatically optimize the execution logic and path selection strategy, and introduce edge computing capabilities to directly process node tasks that require rapid response on edge devices, reduce network latency, and improve execution efficiency. In the process of filling out medical records, the self-learning algorithm optimizes the filling order and prompt information based on the doctor's habits and historical filling data.

[0035] Self-learning algorithm optimizes filling order and prompt information: Suppose a doctor often records the patient's main symptoms first and then the accompanying symptoms when filling out the symptom description part of the medical record. Based on this habit, the self-learning algorithm will prioritize the input box of the main symptoms in the subsequent medical record filling and provide related common symptom options as prompts. For example, when filling out the medical records of a certain disease, doctors often omit some important examination item information. After the self-learning algorithm discovers this problem by analyzing the historical filling data, it will give an obvious prompt in the corresponding position when filling out the medical record of the disease next time, reminding the doctor not to miss it.

[0036] Using edge computing capabilities to quickly process preliminary medical records locally: In the emergency department of a hospital, when a seriously injured patient is sent to the hospital, the doctor needs to record the patient's basic information and preliminary diagnosis as soon as possible. The adaptive execution engine uses edge computing capabilities to quickly process these preliminary medical records directly on the local server of the emergency department without waiting for data interaction with the remote central server. For example, in the case of unstable network, edge computing can ensure that doctors can quickly record the patient's vital signs, injured parts and other key information, and will not delay treatment due to network delays. For another example, in a branch of a hospital or a medical point in a remote area, the network conditions are poor. When there is an emergency medical record to be filled out, edge computing can quickly store and process this data locally, and then synchronize it to the central server after the network returns to normal, ensuring the timely entry of medical record information. Suppose a patient seriously injured in a traffic accident is sent to a branch of a hospital in a mountainous area. The network signal is sometimes available and sometimes not. The adaptive execution engine immediately starts the edge computing mode, and the doctor enters the patient's blood pressure, heart rate, consciousness and other vital signs, as well as the injured part, the general course of the accident and other information on the tablet. These data are quickly stored in the local server, and through a simple algorithm deployed locally, a preliminary judgment is made that the patient may be at risk of internal bleeding, providing key basis for subsequent emergency treatment.

[0037] For emergency cases, the adaptive execution engine module uses edge computing capabilities to quickly process preliminary medical records locally to ensure that information is entered in a timely manner.

[0038] Real-time feedback and dynamic adjustment mechanism module: including real-time feedback unit, including intelligent early warning, using machine learning algorithms to predict potential risks and issue early warning signals in advance, and provide preventive adjustment suggestions; dynamic adjustment mechanism unit: introduce A / B testing mechanism, test different plans in parallel during the dynamic adjustment process, and select the best plan based on data; the intelligent early warning system monitors potential errors in medical records through machine learning algorithms, such as medication conflicts, inconsistent diagnostic logic, etc.

[0039] Real-time feedback unit - intelligent early warning: Medication conflict early warning: Machine learning algorithms learn the interactions and contraindications of different drugs by analyzing a large amount of medical record data. When a doctor prescribes a new prescription for a patient, the algorithm monitors the drug combination in the prescription in real time. If a possible conflict is found, such as prescribing two drugs that may cause serious side effects at the same time, the intelligent early warning system will immediately issue an early warning signal and give preventive adjustment suggestions such as changing one of the drugs or adjusting the dosage.

[0040] Diagnostic logic inconsistency warning: Assume that the patient's symptoms, test results, and preliminary diagnosis are recorded in the medical record. The algorithm can determine the inconsistency between the diagnosis and symptoms or test results by learning the diagnostic patterns and logical relationships of common diseases. For example, if the patient's symptoms and test results strongly point to a certain disease, but the doctor gives a different diagnosis, the system will issue a warning to remind the doctor to review the diagnosis and provide possible correct diagnostic directions as suggestions.

[0041] Dynamic Adjustment Mechanism Unit - A / B Testing: Comparison of Medical Record Templates: Design two different medical record templates, one template focuses on recording the patient's symptoms and signs in detail, and the other template focuses more on treatment plans and follow-up plans. Through the A / B testing mechanism, the medical records of some patients are assigned to use Template A and the other to use Template B. Then compare the data of the two templates in terms of the efficiency of doctors filling in, information completeness, and convenience of subsequent medical teams. If it is found that Template A leads to shorter filling time for doctors and more comprehensive information, Template A will be selected as the main medical record template.

[0042] Filling process optimization: Assume that there are two different medical record filling processes, one is to fill in the patient's medical information in chronological order, and the other is to fill in the key diagnosis information first and then add other details. Through A / B testing, let different doctors use these two processes to fill in medical records. Then evaluate which process is better based on indicators such as filling time, error rate, and convenience of subsequent medical record retrieval, so as to optimize the medical record filling process.

[0043] The real-time feedback and dynamic adjustment mechanism module sends out early warning signals and gives correction suggestions. The A / B testing mechanism is introduced to compare the effects of different medical record templates and filling processes, and optimize medical record management with data-driven.

[0044] Multi-dimensional data analysis module: It expands the functions of the data-driven decision-making optimization system, introduces network analysis tools to reveal the complex relationships between process nodes, and provides comprehensive and in-depth insights and decision-making basis for process optimization. It integrates multi-dimensional data such as the patient's basic information, symptoms, examination results, and treatment plans.

[0045] Data integration: Assume that a hospital's information system includes an electronic medical record system (storing basic patient information and symptom descriptions), a test system (providing test results), and a doctor's order system (including treatment plans). First, extract relevant data from these systems using data extraction tools. For basic patient information, extract name, age, gender, place of residence, etc. For symptoms, obtain detailed symptom descriptions, such as the location, nature, and frequency of pain. Test results include blood indicators, imaging test results, etc. Treatment plans cover the drugs used, treatment methods, treatment cycles, etc. Then, clean and preprocess the extracted data. For example, handle missing values, correct erroneous data, and unify data formats and units. Finally, use data warehouse or data lake technology to integrate and store these processed data to form a unified multi-dimensional patient data set.

[0046] Analyze the disease pattern: Take diabetes as an example, select a large amount of information about diabetic patients from the integrated data set. Analyze the disease incidence of patients of different age groups, for example, find that the incidence rate is higher in the 30-50 age group. Study the relationship between patients' eating habits (such as the frequency of high-sugar and high-fat diets) and the onset of the disease, and you may find that people who often eat high-sugar diets are more likely to get sick. Investigate the impact of family history on the onset of the disease. If there is a diabetic patient in the family, the individual risk of developing the disease is higher.

[0047] Evaluate the effectiveness of treatment: Treatment for a certain type of cancer, such as lung cancer. Compare the survival and recurrence rates of patients under different treatment options (such as surgery, chemotherapy, radiotherapy, or combined therapy). Analyze the side effects of patients during treatment, such as hair loss and nausea caused by chemotherapy, and evaluate the severity of side effects of different drug combinations. Observe the changes in tumor markers after treatment. If the markers decrease significantly, it means that the treatment effect is better. For example, when studying lung cancer treatment options A and B, it was found that option A caused 80% of patients to significantly decrease their tumor markers within 3 months after treatment, while only 60% of patients with option B had this effect, and the side effects of option A were relatively small. Then, in subsequent treatment decisions, option A may be given priority.

[0048] Analysis of disease incidence patterns: Suppose we want to study the incidence patterns of cardiovascular diseases. First, collect multi-dimensional data of patients from multiple data sources such as the hospital's electronic medical record system and health examination database, including age, gender, family medical history, lifestyle habits (such as diet, exercise, smoking and drinking), blood pressure, blood lipids, blood sugar levels, etc. Use data analysis tools and algorithms to preprocess and clean these data to remove noise and missing values. Then, use association rule mining algorithms to find factors that are highly correlated with the incidence of cardiovascular diseases. For example, it was found that people over 50 years old, with a family history of disease, a long-term high-salt and high-fat diet, and lack of exercise have a significantly increased risk of cardiovascular disease. By establishing a prediction model, such as a logistic regression model or a decision tree model, the probability of an individual suffering from cardiovascular disease is predicted based on the above-mentioned related factors. This can help medical institutions intervene and take preventive measures for high-risk groups in advance, such as regular physical examinations, health education, and lifestyle guidance.

[0049] Treatment effect evaluation: Taking cancer treatment as an example, a series of data are collected after patients receive different treatment plans (such as surgery, chemotherapy, radiotherapy, immunotherapy, etc.), including changes in tumor size, survival, quality of life scores, treatment side effects, etc. Survival analysis methods are used to compare the survival rate and survival time of patients under different treatment plans to evaluate their long-term efficacy. At the same time, multifactorial analysis of variance is used to study the effect of the interaction between the treatment plan and factors such as patient age, gender, and cancer stage on the treatment effect. For example, for patients with early lung cancer, it was found that surgical treatment combined with postoperative adjuvant chemotherapy is better than surgical treatment alone; while for patients with advanced lung cancer, immunotherapy combined with chemotherapy may prolong survival. These analysis results can provide a basis for doctors to formulate personalized treatment plans. In addition, by analyzing the data of treatment side effects, such as the incidence and severity of symptoms such as nausea, vomiting, and hair loss, the pros and cons of different treatment plans can be weighed, and the most suitable treatment method can be selected for patients to minimize pain and adverse effects while ensuring efficacy.

[0050] The multi-dimensional data analysis module integrates and deeply mines multi-source medical data, uses advanced data analysis methods and models to discover the incidence patterns and treatment effects of certain diseases, and provides scientific, accurate and valuable information support for medical decision-making, thereby improving medical quality and patient treatment effects.

[0051] Cross-platform and system integration capability module: Provides low-code / no-code integration tools to support non-technical personnel to quickly configure system integration through a graphical interface; supports cloud-native deployment to ensure that the orchestration tools can flexibly adapt to different cloud environments and improve scalability and flexibility. Integrate the electronic medical record system with the hospital's testing system, imaging system, and doctor's order system.

[0052] Implementation of low-code / no-code integration tools: Provide a visual process design interface, allowing non-technical personnel to define the data flow and processing logic between systems through dragging, connecting, and other operations. Built-in rich predefined components and templates, covering common data conversion, validation, and routing rules. Example: Suppose the information department staff of a hospital, who has no programming background, uses a low-code tool to integrate the electronic medical record system with the imaging system. By selecting the icons of the electronic medical record system and the imaging system on the graphical interface, and then dragging the connecting line, set the conditions for data transmission from the imaging system to the electronic medical record system, such as only transmitting specific types of imaging reports. When a new imaging report that meets the conditions is generated, it is automatically transmitted to the electronic medical record system.

[0053] Cloud-native deployment implementation: Use containerization technology to package orchestration tools into containers for rapid deployment and migration in different cloud environments. Use the elastic expansion characteristics of cloud-native to automatically adjust resource allocation according to system load. Example: During peak business periods, such as flu season, the number of visits to the electronic medical record system in hospitals increases significantly. The orchestration tools deployed in cloud-native deployments automatically detect load changes and quickly increase computing and storage resources in the cloud to ensure the responsiveness and stability of the system.

[0054] Implementation using middleware (ESB): Detailed definition of interface specifications between each system and ESB, including data format, communication protocol, etc. Configure data conversion and routing rules on the ESB platform to ensure that data is accurately transmitted between systems. Example: In a large general hospital, the ESB platform receives blood routine data of patients from the testing system, converts the data format into a format recognizable by the electronic medical record system according to preset rules, and accurately routes the data to the corresponding electronic medical record based on the patient's hospitalization department and bed number.

[0055] API-based implementation: Each system develops and publishes a stable and secure API interface, and specifies the input and output parameters and calling methods of the interface. The electronic medical record system calls the interface according to the interface specification and processes the returned data. Example: The doctor's order system of a hospital provides an API interface to obtain information about the patient's allergic drugs. The electronic medical record system calls the interface when the patient visits the doctor, obtains the patient's list of allergic drugs, and reminds the doctor when prescribing to avoid medication errors.

[0056] Implementation method using data warehouse: Design a reasonable data warehouse architecture, including hierarchical storage, index optimization, etc. Develop a job plan for data extraction, cleaning, conversion and loading to ensure the timeliness and accuracy of data. Example: The hospital extracts data from various systems to the data warehouse on a regular basis every day. When the doctor of the electronic medical record system views the patient's medical record details, he can quickly query all the patient's test reports and imaging data for the past year from the data warehouse, providing a comprehensive reference for diagnosis.

[0057] Message queue-based implementation: Each system publishes data or events to the message queue in a unified message format and topic. The electronic medical record system subscribes to topics of interest and processes received messages in a timely manner. Example: After the imaging system completes the patient's MRI examination, it publishes the key information of the examination results to the message queue in the form of a message. The electronic medical record system subscribes to the message in real time and updates the patient's imaging examination status in the medical record, so that the doctor can know in time and make subsequent diagnoses.

[0058] Implementation through microservice architecture: Split each system into independent microservices, and perform service registration and discovery. Define the communication protocol and interface specifications between microservices. Example: The electronic medical record system needs to obtain the latest data provided by the patient's test index microservice. The address and interface of the microservice are discovered through the service registration center. After calling, the latest test index data is obtained and displayed on the medical record page to provide real-time support for the doctor's diagnosis.

[0059] The cross-platform and system integration capability module can effectively realize the integration of the electronic medical record system with other hospital systems, promote the seamless flow and sharing of patient information, and improve medical efficiency and quality. Realize the seamless flow and sharing of patient information and improve medical efficiency.

[0060] Permission management and security audit module: including a permission management unit based on blockchain technology, which records tamper-proof logs of key operations and data changes to improve audit transparency and credibility; a dynamic access control unit that dynamically adjusts access rights based on user behavior and environmental changes. The permission management unit based on blockchain technology records key operations such as creation, modification, and viewing of medical records. The dynamic access control function dynamically adjusts the doctor's access and modification rights to medical records based on the doctor's title and scope of responsibilities.

[0061] Artificial Intelligence Assistant Integration Module: It integrates an intelligent chatbot or virtual assistant unit with adaptive learning capabilities, provides operation guidance and suggestions, and continuously optimizes services based on user habits and feedback, providing doctors with medical record filling standards and medical terminology query services. It continuously optimizes services based on doctors' feedback, such as adding answers to common questions; introduces emotional interaction units to improve user experience and satisfaction.

[0062] Automated testing and verification framework module: An automated testing and verification framework module that includes the CI / CD process was built to ensure that each code change can automatically trigger testing and verification, reducing human errors and online risks. At the same time, chaos engineering experiments were introduced to simulate extreme conditions to test system resilience. Each update to the electronic medical record system can automatically trigger testing and verification to ensure the stability and accuracy of the system.

[0063] Automatically trigger testing and verification: Version control system integration: When developers modify the code of the electronic medical record system and submit it to the version control system (such as Git), the test process is automatically triggered by setting hooks or continuous integration tools (such as Jenkins). Test case library: Establish a comprehensive test case library, including functional test cases (such as creation, modification, query, printing of medical records, etc.), performance test cases (such as response time under high concurrency), compatibility test cases (display and operation on different browsers and operating systems), etc. Automated testing tools: Use automated testing tools (such as Selenium, JMeter, etc.) to execute test cases. For example, Selenium can simulate user operations in the electronic medical record system to verify whether the display and function of interface elements are normal; JMeter can perform stress testing on the system to evaluate the performance of the system when a large number of users access it concurrently. Assuming that the developer updates the code of the medical record printing function in the electronic medical record system, the version control system will perceive this change and automatically trigger the automated testing tool to execute test cases related to the printing function, including printing of medical records in different formats, saving and loading of printing settings, etc., to ensure that the update does not introduce new problems.

[0064] Chaos Engineering Experiments: Fault injection: Use tools (such as ChaosMonkey) to simulate system failures, such as sudden server crashes, database connection interruptions, storage device failures, etc. Network anomaly simulation: Use network simulation tools (such as tc or NetEm) to create network delays, packet loss, bandwidth limitations, etc. Resource limitations: Artificially limit system resources, such as CPU usage, memory usage, disk space, etc., and observe the performance of the system under resource constraints. For example, in order to test the emergency recovery capability of the electronic medical record system in the event of a network interruption, you can use a network simulation tool to suddenly create a network interruption during system operation, restore the network after a period of time, and observe whether the system can automatically reconnect to the database, save local data, and synchronize data after the network is restored to ensure the integrity and accuracy of the medical record data.

[0065] The automated testing and verification framework module can be used in the electronic medical record system to improve the system's stability, reliability and ability to cope with extreme situations. Chaos engineering experiments are introduced to simulate extreme situations such as system failures and network interruptions to test the system's emergency recovery capabilities.

[0066] Comprehensive monitoring and performance optimization module: including comprehensive monitoring unit: real-time monitoring of the entire process of process execution, for performance indicators, resource utilization and user feedback, etc., automatically triggering performance optimization measures to ensure efficient and stable operation of the system. Real-time monitoring of performance indicators of the medical record system, such as response time, storage capacity, etc. When insufficient storage capacity is found, capacity expansion measures are automatically triggered.

[0067] User-defined extension and plug-in support module: It provides user-defined extension interface and plug-in support, allowing users to develop or integrate new functional modules according to business needs, enhancing the flexibility and scalability of the orchestration tool. The hospital has developed a plug-in for docking with the medical insurance system according to its own needs to facilitate the transmission of medical insurance reimbursement data. Departments are allowed to develop functional modules according to special medical record management requirements, such as follow-up record modules for specific diseases.

[0068] Through the application of the above-mentioned intelligent adaptive process node orchestration tool in the electronic medical record management of the medical industry, the intelligent identification unit and the dynamic construction unit can also work closely with other modules and combine with the adaptive execution engine module to provide it with accurate process architecture information so that it can execute efficiently; interact with the multi-dimensional data analysis module to obtain data support to optimize the construction strategy; work with the authority management and security audit module to ensure that the constructed process complies with security specifications and authority requirements; improve the quality and efficiency of medical record management, ensure the security and effective use of medical data, and provide patients with better medical services. Embodiment 2

[0069] A process node-based orchestration tool used in enterprise human resource management systems, such as the optimization and management of employee recruitment and training processes.

[0070] See also Figure 1 , the present invention provides a technical solution:

[0071] See also Figure 1 As shown in the figure, a process node-based orchestration tool includes intelligent identification and dynamic construction modules: the module can automatically identify recruitment needs for different positions, such as technical R&D positions, marketing positions, etc. According to the characteristics of the position and the company's recruitment strategy, the recruitment process is dynamically constructed, including the selection of channels for publishing recruitment information, the standards and processes for screening resumes, etc.

[0072] Adaptive execution engine module: During the recruitment process, the engine's integrated self-learning algorithm continuously collects candidate feedback and interview performance data. It automatically optimizes the execution logic of the recruitment process, such as adjusting interview rounds and focus. For urgent recruitment needs that need to be processed quickly, edge computing capabilities are used to directly process related tasks on local devices, reducing network latency and improving response speed.

[0073] Real-time feedback and dynamic adjustment mechanism module: The intelligent early warning system in the real-time feedback system uses machine learning algorithms to monitor potential risks in the recruitment process, such as delayed recruitment progress, to issue early warning signals in advance, and to provide adjustment suggestions to speed up the recruitment process. Introduce an A / B testing mechanism to test different recruitment channels and screening methods in parallel, and select the most effective recruitment plan driven by data.

[0074] Multi-dimensional data analysis module: Integrate various data of candidates, such as education background, work experience, skill certificates, etc. Analyze the data to find that candidates with certain educational backgrounds perform better in specific positions, providing reference for subsequent recruitment.

[0075] Cross-platform and system integration capability module: seamlessly integrate the recruitment system with the company's internal human resources management system, performance management system, etc. to achieve data circulation and collaboration, such as automatically synchronizing the information of successfully recruited employees to the human resources management system.

[0076] Permission management and security audit module: Built-in permission management unit based on blockchain technology, recording key operations and data changes in the recruitment process, such as interviewer evaluation and hiring decisions. Dynamic access control function dynamically adjusts employees' access rights to recruitment data according to their positions and responsibilities.

[0077] AI assistant integration module: Integrate intelligent chatbots to provide operational guidance to recruiters, such as how to write attractive recruitment copy. Continuously optimize answers and suggestions based on user habits and feedback. Introduce emotional interaction functions to provide encouragement and support when recruiters are under great pressure.

[0078] Automated testing and verification framework module: Build an automated testing and verification framework unit that includes CI / CD processes. Each code change to the recruitment system can automatically trigger testing and verification. Introduce chaos engineering experiments to simulate extreme conditions such as system failures and network delays to test the stability and resilience of the recruitment system.

[0079] Comprehensive monitoring and performance optimization module: The comprehensive monitoring system monitors the execution of the recruitment process in real time, including recruitment progress, resource input, candidate satisfaction, etc. When it is found that the utilization rate of recruitment resources is insufficient, optimization measures are automatically triggered, such as adjusting the division of labor of the recruitment team.

[0080] User-defined extension and plug-in support module: The company developed an integrated plug-in with the external talent pool system based on its own business needs, expanding recruitment channels. The human resources department is allowed to develop new functional modules based on business changes, such as assessment tools for specific positions. Through the application of the above intelligent adaptive process node orchestration tools in the enterprise human resources management system, the recruitment and training process has been made highly intelligent, humanized and secure, which has improved work efficiency and effectiveness, and provided strong talent support for the development of the enterprise. Embodiment 3

[0081] A process node-based orchestration tool used in automobile manufacturing plants. The production process involves multiple links such as parts procurement, assembly, and quality inspection.

[0082] See also Figure 1 , the present invention provides a technical solution:

[0083] See also Figure 1 As shown in the figure, a process node-based orchestration tool includes intelligent recognition and dynamic construction modules: the intelligent recognition unit can extract key features such as vehicle model and configuration from order information, and the dynamic construction unit generates an optimized production process architecture in real time based on these features and rules such as production capacity and inventory status. For example, when a large number of orders for a specific vehicle model are received, the layout of the production line and the staffing arrangement are automatically adjusted.

[0084] Adaptive execution engine module: For urgent quality inspection tasks on the production line, if defects are found in key components, the adaptive execution engine uses edge computing capabilities to quickly process inspection data on edge devices at the production site and promptly decide whether to suspend the production line to avoid the generation of a large number of defective products.

[0085] Real-time feedback and dynamic adjustment mechanism module: The real-time feedback unit can collect data such as the equipment operation status and worker operation efficiency on the production line in real time. The dynamic adjustment mechanism unit introduces an A / B testing mechanism, such as testing two different parts assembly sequences, to select a more efficient and better quality solution driven by data.

[0086] Multi-dimensional data analysis module: Use network analysis tools to reveal the complex relationships between various links in the production process. For example, discover the impact of a delay in the supply of a certain component on the entire production schedule, and provide a decision-making basis for optimizing the supply chain.

[0087] Cross-platform and system integration capability module: Integrate the production management system with the supply chain system, quality inspection system, sales system, etc. For example, when the sales system receives special customization requirements from customers, the relevant information can be quickly transmitted to the production management system to adjust the production process.

[0088] Permission management and security audit module: The permission management unit based on blockchain technology ensures that only authorized personnel can change production plans and key process parameters, while recording tamper-proof logs of these operations. The dynamic access control unit dynamically adjusts the access rights of employees to production data according to job changes and project requirements.

[0089] Artificial intelligence assistant integration module: The intelligent chat robot unit can answer workers' questions about production processes and operating specifications. The emotional interaction unit provides reference for optimizing the working environment and management methods by analyzing employees' emotional feedback at work.

[0090] Automated testing and verification framework module: Introduces chaos engineering experiments to simulate extreme conditions such as equipment failures and raw material shortages to test the resilience and emergency response capabilities of the production system.

[0091] Comprehensive monitoring and performance optimization module: The comprehensive monitoring unit monitors the equipment performance indicators, parts inventory utilization and customer feedback on the product on the production line in real time. When the failure rate of a certain equipment increases or the inventory turnover rate decreases, performance optimization measures are automatically triggered to ensure efficient and stable operation of the system, such as arranging equipment maintenance or adjusting procurement plans.

[0092] User-defined extension and plug-in support modules: Factories can develop customized functional plug-ins based on their own special production processes and management needs, such as a processing flow control module for a new type of material, and integrate it into the orchestration tool.

[0093] In summary: The present invention integrates different modules to create a comprehensive and highly comprehensive process node orchestration tool. The intelligent identification and dynamic construction, adaptive execution engine, real-time feedback and dynamic adjustment mechanism, multi-dimensional data analysis and other modules are innovatively integrated, successfully breaking through the bottleneck of single function and isolation of traditional tools. Cutting-edge technologies and methods are introduced in each module. For example, self-learning algorithms and edge computing capabilities are integrated in the adaptive execution engine, machine learning algorithms are used for intelligent early warning and A / B testing mechanisms are introduced in real-time feedback and dynamic adjustment mechanisms, and blockchain technology is used in authority management and security audit modules. It has significant foresight. The design concept of the entire tool is closely centered on flexibility and scalability. With the help of user-defined extensions and plug-in support modules, users are fully empowered with the ability to carry out personalized development and integration according to their own business needs, thereby providing broad application possibilities for different industries and diversified scenarios. It has high technical value and broad application prospects.

Claims

1. A process node-based orchestration tool used to optimize and manage the employee recruitment process in the enterprise human resource management system. Features: Includes the following modules: Intelligent identification and dynamic construction module: including: Intelligent identification unit: extracts key features and elements from input process information, and the module can automatically identify recruitment needs for different positions; Dynamic construction unit: can generate flexible and optimized process architecture in real time based on the identified process information and specific rules and strategies, dynamically adjust and build process nodes, and dynamically build recruitment processes based on job characteristics and the company's recruitment strategy, including the selection of channels for publishing recruitment information, and the standards and processes for screening resumes; Adaptive execution engine module: Introducing edge computing capabilities to directly process node tasks that require quick response on edge devices. For urgent recruitment needs that need to be processed quickly, edge computing capabilities are used to directly process related tasks on local devices, reducing network delays and improving execution efficiency. Real-time feedback and dynamic adjustment mechanism module: including real-time feedback unit: the intelligent early warning system in the real-time feedback system monitors the potential risks in the recruitment process through machine learning algorithms, and issues early warning signals for lagging recruitment progress; dynamic adjustment mechanism unit: introduces A / B testing mechanism, tests different solutions in parallel during the dynamic adjustment process, and selects the best solution driven by data; Multi-dimensional data analysis module: Integrate various data of candidates, including educational background, work experience, and skill certificates. Analyze the data to find that candidates with certain educational backgrounds perform better in specific positions, providing reference for subsequent recruitment; Cross-platform and system integration capability module: It provides low-code / no-code integration tools to support non-technical personnel to quickly configure system integration through a graphical interface. It also supports cloud-native deployment to ensure that the orchestration tools can flexibly adapt to different cloud environments and improve scalability and flexibility. It also seamlessly integrates the recruitment system with the company's internal human resources management system and performance management system to achieve data circulation and collaboration. Permission management and security audit module: including a permission management unit based on blockchain technology, which records key operations and data changes in the recruitment process; dynamic access control function dynamically adjusts employees' access rights to recruitment data according to their positions and responsibilities; AI assistant integration module: including integrated intelligent chatbot to provide operational guidance to recruiters; emotional interaction unit: to provide encouragement and support when recruiters are under great pressure; Automated testing and verification framework module: Introduce chaos engineering experiments to simulate extreme conditions to test system resilience: simulate system failures and extreme network delay conditions to test the stability and resilience of the recruitment system.

2. A process node-based orchestration tool as claimed in claim 1, Features: The multi-dimensional data analysis module: expands the functionality of the data-driven decision optimization system, introduces network analysis tools to reveal the complex relationships between process nodes, and provides comprehensive and in-depth insights and decision-making basis for process optimization.

3. A process node-based orchestration tool as claimed in claim 1, Features: The automated testing and verification framework module builds an automated testing and verification framework unit that includes a CI / CD process to ensure that each code change can automatically trigger testing and verification.

4. A process node-based orchestration tool as claimed in claim 1, Features: It also includes comprehensive monitoring and performance optimization modules: including comprehensive monitoring unit: real-time monitoring of the entire process of process execution, used for performance indicators, resource utilization and user feedback, and automatically triggering performance optimization measures.

5. The process node-based orchestration tool according to claim 1, Features: It also includes user-defined extension and plug-in support modules: it provides user-defined extension interfaces and plug-in support, allowing users to develop or integrate new functional modules according to business needs, thereby enhancing the flexibility and scalability of the orchestration tool.

Citation Information

Patent Citations

  • Artificial intelligence management system based on cross-platform data interaction

    CN117271981A

  • Office cooperation system and method based on artificial intelligence

    CN118521272A