Workflow arrangement method and device and electronic equipment
The call order of AI capabilities is dynamically adjusted through the workflow orchestration method, which solves the problem of poor flexibility in AI capabilities application in multimedia calls, and achieves more efficient application of AI capabilities.
Patent Information
- Application Number
- CN202411098494.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-12
- Publication Date
- 2024-11-29
Smart Images

Figure CN119052215A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of Internet technology, and in particular to a workflow arrangement method, device, electronic device, storage medium and program product. Background Art
[0002] With the commercialization of the fifth generation wireless systems (5G) technology, multimedia calls are no longer simple voice exchanges, but integrate multiple media forms such as video, voice, text, and images. In this context, the integration of artificial intelligence (AI) technology, especially AI capabilities such as natural language processing, computer vision, speech recognition, and speech synthesis, has further enhanced the interactivity and fun of multimedia calls. For example, the implementation of real-time translation services often requires connecting multiple capability modules in series, such as combining AI capabilities such as speech recognition, speech transcription, speech translation, and speech synthesis to form a complete translation process.
[0003] In some scenarios, the combination of AI capabilities for multimedia calls relies on hard coding, that is, developers need to customize each AI capability in a fixed order according to specific needs. Although this method can achieve a high degree of customization when processing specific services, its disadvantages are also very obvious. The protocols of each AI capability are different, and the combination and calling order of AI capabilities are fixed, and cannot be dynamically adjusted according to different scenarios and needs. If there are new services, it is necessary to redevelop a set of codes to adapt to the new AI capability combination call. This means that every business change requires a lot of manpower and time to modify and test the combination of AI capabilities, and the application of AI capabilities is inflexible and inefficient. Summary of the invention
[0004] The purpose of the embodiments of the present application is to provide a workflow orchestration method, device, electronic device and storage medium, which solve the problem of poor application flexibility and low efficiency of AI capabilities.
[0005] In order to solve the above technical problems, the embodiments of the present application are implemented as follows:
[0006] In a first aspect, an embodiment of the present application provides a workflow orchestration method, which includes: obtaining a target voice service for a multimedia call; in response to a first input from a user, selecting a first target capability corresponding to the target voice service from an artificial intelligence capability library; arranging a calling sequence of the first target capability based on the voice service requirements of the target voice service to obtain a workflow; and completing the target voice service by calling the first target capability according to the calling sequence in the workflow.
[0007] In a second aspect, an embodiment of the present application provides a workflow orchestration device, which includes: an acquisition module for acquiring a target voice service for a multimedia call; a selection module for selecting a first target capability corresponding to the target voice service from an artificial intelligence capability library in response to a first input from a user; an orchestration module for orchestrating a calling sequence of the first target capability based on the voice service requirements of the target voice service to obtain a workflow; and a calling module for completing the target voice service by calling the first target capability according to the calling sequence in the workflow.
[0008] In a third aspect, an embodiment of the present application provides an electronic device, comprising a processor, a communication interface, a memory and a communication bus; wherein the processor, the communication interface and the memory communicate with each other via the communication bus; the memory is used to store computer programs; the processor is used to execute the programs stored in the memory to implement the steps of the workflow orchestration method mentioned in the first aspect.
[0009] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the workflow orchestration method mentioned in the first aspect are implemented.
[0010] In a fifth aspect, an embodiment of the present application provides a computer program product, which includes a computer program stored on a non-transitory computer-readable storage medium, and the computer program includes program instructions. When the program instructions are executed by a computer, the computer executes the steps of the workflow orchestration method mentioned in the first aspect.
[0011] Through the technical solution disclosed in the embodiment of the present application, the target voice service of the multimedia call is obtained; then in response to the first input of the user, the first target capability corresponding to the target voice service is selected from the capability library of the artificial intelligence; the calling sequence of the first target capability is arranged based on the voice service requirements of the target voice service to obtain a workflow; finally, the target voice service is completed by calling the first target capability according to the calling sequence in the workflow. In this way, specific capabilities can be selected according to the actual needs of the target voice service, and then the calling sequence of each capability can be arranged based on the actual needs. The capabilities can be dynamically combined and the calling sequence can be arranged according to different scenarios and needs, thereby improving the configurability of the capabilities of artificial intelligence, improving the application flexibility and efficiency of the capabilities of artificial intelligence, and meeting diversified and personalized business needs. BRIEF DESCRIPTION OF THE DRAWINGS
[0012] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.
[0013] Figure 1 A flowchart of a workflow arrangement method provided in an embodiment of the present application;
[0014] Figure 2 A flowchart of serial arrangement provided for an embodiment of the present application;
[0015] Figure 3 A flowchart of parallel arrangement provided for an embodiment of the present application;
[0016] Figure 4 A flowchart of the conditions provided in the embodiments of the present application;
[0017] Figure 5 A schematic diagram of the process of orchestration and debugging provided in an embodiment of the present application;
[0018] Figure 6 A schematic diagram of the module composition of a workflow arrangement device provided in an embodiment of the present application;
[0019] Figure 7 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION
[0020] The purpose of the embodiments of the present application is to provide a workflow orchestration method, device, electronic device and storage medium, which solve the problem of poor application flexibility and low efficiency of AI capabilities.
[0021] In order to enable those skilled in the art to better understand the technical solutions in this application, the technical solutions in the embodiments of this application will be clearly and completely described below in conjunction with the drawings in the embodiments of this application. Obviously, the described embodiments are only part of the embodiments of this application, not all of them. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of this application.
[0022] For example, Figure 1 As shown, the embodiment of the present application provides a workflow orchestration method, the execution subject of the method may be a server, and the workflow orchestration method may specifically include the following steps:
[0023] In step S101, a target voice service of a multimedia call is obtained.
[0024] Specifically, the multimedia call may be a multimedia call conducted in a 5G network environment, which may support real-time transmission and interaction of audio, video and other media content. The target voice service may be a service in a real-time scenario of an audio or video call, such as an intelligent translation service, an interesting call service and an intelligent customer service.
[0025] In step S103, in response to a first input from a user, a first target capability corresponding to a target voice service is selected from a capability library of artificial intelligence.
[0026] Specifically, various capabilities of the AI capability library can be configured in the orchestration system, and the page of the orchestration system can display various AI capabilities. Users can select the capabilities actually required by the target voice service on the page of the orchestration system. For example, if the target voice service is an intelligent translation service, users can select the three capabilities of voice recognition, voice transcription, and voice translation on the page of the orchestration system. Specifically, the selection can be performed by the user by clicking, selecting a box, long pressing, etc., and selecting the AI capability displayed on the page of the orchestration system as the first target capability selected.
[0027] In step S105, the calling sequence of the first target capability is arranged based on the voice service demand of the target voice service to obtain a workflow.
[0028] Specifically, after selecting the first target capability, the calling order of the first target capability can be arranged according to the voice service requirements. The calling order can be sequential order, parallel order, etc. For example, the target voice service is an intelligent translation service, and the business requirements of the intelligent translation service are voice recognition, transcription, and translation. The user can select the three capabilities of voice recognition, voice transcription, and voice translation on the page of the orchestration system. These three capabilities have a strict sequence in application, that is, recognition first, then transcription, and finally translation. According to this order, the calling sequence of the three capabilities of voice recognition, voice transcription, and voice translation is arranged.
[0029] When arranging the first target capability, multiple methods may be used. The calling sequence of the first target capability is arranged based on the voice service requirements of the target voice service, and the obtained workflow includes:
[0030] Determine an orchestration mode of the first target capability based on the voice service demand of the target voice service; orchestrate the calling sequence of the first target capability according to the orchestration mode to obtain a workflow, and the orchestration modes include: group orchestration, serial orchestration, parallel orchestration and conditional orchestration; serial orchestration indicates that each first target capability is called in sequence, parallel orchestration indicates that each first target capability is called in parallel, conditional orchestration indicates that the first target capability is called according to a preset condition, and group orchestration indicates that the first target capability is called in groups.
[0031] Specifically, for different voice service needs, the execution order of the first target capabilities that meet them can be executed in sequence or in parallel. Therefore, it is necessary to determine the arrangement method of the call order of each first target capability according to the voice service needs. For example, if the target voice service is an intelligent translation service, and the business requirements of the intelligent translation service are voice recognition, transcription, and translation, the user can select the three capabilities of voice recognition, voice transcription, and voice translation on the page of the arrangement system. These three capabilities have a strict order of precedence when applied, that is, they need to be recognized first, then transcribed, and finally translated. According to this order, the order of calling the three capabilities of voice recognition, voice transcription, and voice translation is serially arranged. The target voice service is a fun call service, and the business requirements of the fun call are voice changing service and beauty service. The user can select the two capabilities of voice synthesis and beauty on the page of the arrangement system. These two capabilities can be executed in parallel when applied, that is, there is no order of calling. The user can choose to arrange the calling order of voice synthesis capability and beauty capability in a parallel arrangement manner, that is, voice synthesis capability and beauty capability can be executed at the same time.
[0032] Further, such as Figure 2 As shown, serial arrangement can realize the orderly calling of multiple artificial intelligence capabilities, so as to realize more complex business logic, and Task 1 to Task 3 are arranged in order. For example, you can first call an AI capability for speech recognition, then call another AI capability for semantic analysis, and finally call the third AI capability for natural language generation, so that more intelligent multimedia calls can be realized.
[0033] like Figure 3 As shown, parallel orchestration enables AI capabilities to be executed simultaneously. Figure 3In the example, there is a sequence between Task 1 and Task 2, and Task 2 corresponds to parallel Task 3 to Task 5, which can be executed in parallel. In actual scenarios, parallel orchestration is a more efficient method that allows multiple AI capabilities to execute simultaneously and interact and integrate their outputs. The advantage of this is that multiple tasks can be processed at the same time, thereby improving efficiency and speed, while also providing more accurate and comprehensive results by more comprehensively considering the relationships between different tasks. Furthermore, using parallel computing and asynchronous processing, the calls to AI capabilities are distributed to multiple threads or processes to achieve parallel calls and asynchronous processing. This can also improve operating efficiency and resource utilization, and reduce response time and error rate.
[0034] like Figure 4 As shown in the figure, conditional orchestration is a technology that combines and orchestrates AI capabilities in a way that determines the subsequent branch processes based on the pre-parameter conditions. By conditionally judging the input parameters to determine which branch process to execute, more flexible and customizable AI applications can be achieved. In traditional orchestration methods, the process of AI capabilities is linear and fixed, and cannot be dynamically adjusted according to different parameter conditions. Conditional orchestration has greater flexibility in this regard. It allows AI capabilities to choose to execute different branch processes based on different pre-parameter conditions to meet different needs and scenarios.
[0035] The grouping arrangement indicates that the first target capabilities are called in groups, wherein the first target capabilities with high correlation can be divided into a group, and the first target capabilities in the group are called in parallel, and the AI capabilities between the groups can be called sequentially or in parallel according to actual needs. Alternatively, the first target capabilities with high similarity can be divided into a group, and the first target capabilities in the group are called in parallel, and the AI capabilities between the groups can be called sequentially or in parallel according to actual needs.
[0036] Specifically, the pipeline relevance algorithm can be used to group various AI capabilities. According to the similarity or relevance between AI capabilities, AI capabilities with high similarity or relevance can be grouped into the same group for orchestration and execution, while AI capabilities with low similarity or relevance can be separated and executed in a staggered manner to improve overall execution efficiency and resource utilization. When using the relevance algorithm to calculate the similarity or relevance between AI capabilities, the similarity or relevance between AI capabilities can be calculated based on factors such as the input-output relationship, dependency, and data flow between AI capabilities.
[0037] The efficiency of a pipeline refers to the utilization rate of the AI capabilities of the pipeline. In the time-space diagram, the efficiency E (correlation or similarity) of a pipeline is defined as the ratio of the time-space area occupied by n tasks to the total time-space area of k pipeline segments, as shown in the following formula:
[0038]
[0039] Based on the above correlation algorithm, we can build a correlation matrix or graph between AI capabilities, and divide AI capabilities into multiple groups through clustering, graph segmentation, etc. AI capabilities with high similarity or correlation will be divided into the same group, while AI capabilities with low similarity or correlation will be divided into different groups, so as to realize grouped orchestration and execution of AI capabilities.
[0040] In the process of grouping and arranging the execution of AI capabilities, AI capabilities in the same group can be executed in parallel, thereby improving the overall execution efficiency. Because there are similarities or correlations between these AI capabilities, their execution order and dependencies are relatively close. However, AI capabilities in different groups can be executed in a staggered manner due to their low similarity or correlation, making full use of system resources and optimizing execution efficiency.
[0041] The implementation of AI capability grouping technology helps improve the performance and resource utilization of the orchestration system. It reasonably divides AI capability groups to reduce the dependencies and redundant calculations between AI capabilities and achieve more efficient parallel execution. At the same time, AI capability grouping can also help users better understand and manage the orchestration process, thereby improving the efficiency of development and maintenance.
[0042] In summary, the grouping technology uses the pipeline correlation algorithm to group AI capabilities, and puts AI capabilities with high similarity or correlation in the same group for orchestration and execution, thereby improving overall execution efficiency and resource utilization. This technology can calculate the similarity of AI capabilities through correlation algorithms, build correlation matrices or graphs, and use clustering, graph segmentation and other methods to group AI capabilities. Through the pipeline correlation algorithm, the AI capabilities in the workflow are grouped and associated to optimize execution efficiency, reasonably allocate computing resources, and improve resource utilization. These AI capability combinations can be optimized through the pipeline correlation algorithm to achieve the best execution order and resource allocation, avoiding resource waste and performance bottlenecks.
[0043] Furthermore, when orchestrating AI capabilities, you can also perform orchestration debugging, specifically by initiating a debugging request to the browser front-end page; calling the inference platform back-end interface through the browser front-end page; calling the orchestration service through the inference platform back-end interface to execute a custom script to debug the orchestration environment. Figure 5As shown, the user can initiate a debugging request to the browser front-end page by writing a script, and then the browser front-end page calls the inference platform back-end interface (which can be a script and set parameters), and the inference platform back-end interface calls the orchestration service to execute the custom script. When the orchestration service executes the custom script, it generates a Py file with the pre-defined python code and the user-set parameters and executes it to debug the orchestration environment. Finally, the orchestration service returns the debugging results to the inference platform back-end, and finally returns to the browser front-end page to display the debugging results of the custom script on the orchestration environment. Among them, the debugging part is not limited to the following three types:
[0044] The first is the media file upload test of multimedia calls.
[0045] During the orchestration debugging process, we first need to test the upload function of media files such as audio and video of multimedia calls, select some typical media files, and upload them to the orchestration environment. Ensure that the uploaded file format is consistent with the requirements of the system in the orchestration environment, and verify the correctness and stability of the file upload. This can be automated by writing test cases and scripts to ensure the effective upload of media files.
[0046] Then the intermediate results are displayed.
[0047] In order to clearly observe the intermediate parameter transfer process of the orchestration of AI capabilities, it is necessary to add intermediate nodes in the orchestration system and display the intermediate results in these nodes. You can use visual editing tools to create nodes to display specific variables, data, or intermediate states, so that you can easily observe and debug intermediate results to ensure correct parameter transfer and processing.
[0048] Finally, logging and monitoring are performed.
[0049] During the orchestration debugging process, by recording the operation logs of the system in the orchestration environment, we can track and analyze the execution of the orchestration process and the transmission process of intermediate parameters, which can help us locate and solve problems. In addition, the system is also equipped with a suitable monitoring system to monitor the performance and operation status of the system and promptly discover and solve potential problems.
[0050] In step S107, the target voice service is completed according to the first target capability called in the calling sequence in the workflow.
[0051] Specifically, after the workflow is arranged, the first target capability is called according to the calling sequence in the workflow to sequentially execute various tasks in the target voice service to complete the target voice service.
[0052] More specifically, for the target voice service, media stream pre-processing is required. Media stream pre-processing is a kind of encoding processing of media streams. The main purpose is to convert the original audio and video data into a media stream suitable for transmission, and transmit it to the AI capability for subsequent processing and analysis to achieve more efficient and customizable audio and video applications. First, the audio and video data needs to be encoded and compressed into a format suitable for transmission and storage, such as H.264, AAC, etc. The encoded media stream can greatly reduce the amount of data, improve transmission efficiency, and maintain good audio and video quality. Next, the encoded media stream can be transmitted to the AI capability, and the encoded media stream can be further processed and analyzed by the AI capability. For example, AI capabilities can also include functions such as speech recognition, image recognition, and video analysis. By passing the encoded media stream to the AI capability, tasks such as deep learning, semantic understanding, and sentiment analysis of audio and video content can be achieved.
[0053] Through the technical solution disclosed in the embodiment of the present application, the target voice service of the multimedia call is obtained; then in response to the first input of the user, the first target capability corresponding to the target voice service is selected from the capability library of the artificial intelligence; the calling sequence of the first target capability is arranged based on the voice service requirements of the target voice service to obtain a workflow; finally, the target voice service is completed by calling the first target capability according to the calling sequence in the workflow. In this way, specific capabilities can be selected according to the actual needs of the target voice service, and then the calling sequence of each capability can be arranged based on the actual needs. The capabilities can be dynamically combined and the calling sequence can be arranged according to different scenarios and needs, thereby improving the configurability of the capabilities of artificial intelligence, improving the application flexibility and efficiency of the capabilities of artificial intelligence, and meeting diversified and personalized business needs.
[0054] Before selecting a first target capability corresponding to a target voice service from a capability library of an artificial intelligence in response to a first input from a user, the method further includes: registering each capability of the artificial intelligence in the capability library of the artificial intelligence.
[0055] Specifically, AI capability registration can be a prerequisite for AI capability orchestration. The goal of AI capability registration is to pre-define and register AI capabilities in the orchestration system so that these AI capabilities can be called and managed in subsequent communication processes. AI capability registration includes Hypertext Transfer Protocol (HTTP) registration, Transmission Control Protocol-based full-duplex communication protocol (WebSocket) registration, and JSON message parsing function.
[0056] Among them, registering each capability of artificial intelligence in the capability library of artificial intelligence includes: registering each capability in the capability library of artificial intelligence based on the hypertext transfer protocol.
[0057] Specifically, HTTP protocol capability registration is one of the key parts of AI capability registration. HTTP is a stateless application layer protocol based on TCP / IP. It is responsible for transmitting hypertext on the network and providing users with a request-response working mode. In the registration of AI capabilities, the characteristics of the HTTP protocol are used to realize the registration of AI capabilities. More specifically, the request header, request body, request parameters, response header, response body, and response code of the HTTP protocol are registered separately, and the registration of the AI capability of the HTTP protocol is completed by defining each protocol segment.
[0058] Furthermore, registering each capability of the artificial intelligence in a capability library of the artificial intelligence includes: registering each capability in the capability library of the artificial intelligence based on a full-duplex communication protocol of the transmission control protocol.
[0059] Specifically, WebSocket protocol capability registration is another key part of AI capability registration. WebSocket is a protocol for full-duplex communication over a single Transmission Control Protocol (TCP) connection. WebSocket makes data exchange between clients and servers easier, allowing the server to actively push data to the client. In AI capability registration, we use the characteristics of the WebSocket protocol to achieve real-time, two-way, and persistent AI capability registration. More specifically, the client messages and server messages of the WebSocket protocol are registered separately, and the message format is generated by defining the message body to facilitate message delivery during the orchestration process.
[0060] In one possible implementation, after the calling sequence of the first target capability is arranged based on the voice service demand of the target voice service to obtain the workflow, the method also includes: in response to the second input of the user, selecting the second target capability from the capability library of the artificial intelligence; and arranging the first target capability based on the voice service demand of the target voice service to obtain the workflow, which includes: arranging the calling sequence of the first target capability and the second target capability based on the voice service demand of the target voice service to obtain the workflow.
[0061] Specifically, after the user arranges the workflow of the target voice service, the user can also add other AI capabilities to the workflow. The user can further click, long press or box on the page of the arrangement system to select the second target capability, which is different from the first target capability, and then re-arrange the first target capability and the second target capability according to the needs of the voice service to obtain a new workflow. In this way, the user can flexibly reconfigure the workflow according to actual needs, which is highly flexible and improves the user experience.
[0062] In a possible implementation, after arranging the calling order of the first target capabilities based on the voice service requirements of the target voice service to obtain the workflow, the method further includes: adjusting the calling order of each first target capability in the workflow in response to a third input from the user.
[0063] Specifically, the third input can be dragging, clicking, box selection, etc. For example, the user can drag the ability whose position they want to adjust in the workflow and drag it to the position of another ability, and the positions of the two abilities will be exchanged.
[0064] In a possible implementation, after arranging the calling sequence of the first target capabilities based on the voice service requirements of the target voice service to obtain the workflow, the method further includes: deleting part of the first target capabilities in the workflow in response to a fourth input from the user.
[0065] Specifically, the fourth input can be dragging, clicking, box selection, etc. For example, when a user wants to delete a capability in a workload, he can click on the capability and then select delete in the pop-up toolbar. In this way, the user can flexibly reconfigure the workflow according to actual needs, which is highly flexible and improves the user experience.
[0066] It is worth noting that the orchestration system provides a front-end page through which users can flexibly configure AI capabilities.
[0067] In addition, you can also set up custom nodes, which process special logic by writing code online. Custom code nodes can be added to the orchestration system to implement specific functions or meet predetermined requirements, thereby achieving higher freedom and flexibility. Traditional orchestration systems usually provide some basic nodes and functions, such as data conversion, conditional judgment, loop control, etc., for building orchestration processes. However, sometimes we may need to implement some complex logic or special processing requirements, which these basic nodes cannot meet. At this time, custom nodes play a key role. Through custom nodes, we can write code online and use various programming languages and technologies to process specific logic or implement custom functions. This makes the functions of the orchestration system richer and more extended, and can flexibly respond to various scenarios and needs.
[0068] Corresponding to the workflow arrangement method provided in the above embodiment, based on the same technical concept, the embodiment of the present application also provides a workflow arrangement device, Figure 6 A schematic diagram of the module composition of a workflow arrangement device provided in an embodiment of the present application, wherein the workflow arrangement device is used to execute Figure 1 Describes the workflow orchestration method, such as Figure 6 As shown, the workflow orchestration device 600 includes: an acquisition module 601, which is used to acquire the target voice service of the multimedia call; a selection module 602, which is used to select a first target capability corresponding to the target voice service from the capability library of the artificial intelligence in response to the first input of the user; an orchestration module 603, which is used to orchestrate the calling sequence of the first target capability based on the voice service demand of the target voice service to obtain the workflow; and a calling module 604, which is used to complete the target voice service by calling the first target capability according to the calling sequence in the workflow.
[0069] It can be seen from the technical solution provided by the above embodiment of the present application that the target voice service of the multimedia call is obtained; then in response to the first input of the user, the first target capability corresponding to the target voice service is selected from the capability library of the artificial intelligence; the calling sequence of the first target capability is arranged based on the voice service requirements of the target voice service to obtain a workflow; finally, the first target capability called according to the calling sequence in the workflow completes the target voice service. In this way, specific capabilities can be selected according to the actual needs of the target voice service, and then the calling sequence of each capability can be arranged based on the actual needs. The capabilities can be dynamically combined and the calling sequence can be arranged according to different scenarios and needs, thereby improving the configurability of the capabilities of artificial intelligence, improving the application flexibility and efficiency of the capabilities of artificial intelligence, and meeting diversified and personalized business needs.
[0070] In a possible implementation, it also includes: a registration module, which is used to register various capabilities of artificial intelligence in a capability library of artificial intelligence.
[0071] In a possible implementation, the registration module is also used to register each capability in the capability library of the artificial intelligence based on the hypertext transfer protocol.
[0072] In a possible implementation, the registration module is also used to register each capability in the capability library of the artificial intelligence based on the full-duplex communication protocol of the transmission control protocol.
[0073] In a possible implementation, the selection module 602 is further used to select a second target capability from the capability library of the artificial intelligence in response to a second input from the user; the orchestration module 603 is further used to orchestrate the calling sequence of the first target capability and the second target capability based on the voice service requirements of the target voice service to obtain a workflow.
[0074] In a possible implementation, the orchestration module 603 is further configured to adjust the order of calling the first target capabilities in the workflow in response to a third input from the user.
[0075] In a possible implementation, the orchestration module 603 is further configured to delete part of the first target capabilities in the workflow in response to a fourth input from the user.
[0076] In a possible implementation, the orchestration module 603 is further used to determine an orchestration mode of the first target capability based on the voice service demand of the target voice service; the calling order of the first target capability is orchestrated according to the orchestration mode to obtain a workflow, and the orchestration modes include: group orchestration, serial orchestration, parallel orchestration and conditional orchestration; the serial orchestration indicates that each first target capability is called in sequence, the parallel orchestration indicates that each first target capability is called in parallel, the conditional orchestration indicates that the first target capability is called according to a preset condition, and the group orchestration indicates that the first target capability is called in groups.
[0077] In a possible implementation manner, the grouping includes: grouping the first target capabilities with high correlation into a group, and calling the first target capabilities in the group in parallel.
[0078] In a possible implementation, before the calling sequence of the first target capability is orchestrated based on the voice service requirements of the target voice service to obtain the workflow, the method also includes: initiating a debugging request to the browser front-end page; calling the reasoning platform back-end interface through the browser front-end page; calling the orchestration service through the reasoning platform back-end interface to execute a custom script to debug the orchestration environment.
[0079] The workflow orchestration device provided in the embodiment of the present application can implement each process in the embodiment corresponding to the above-mentioned workflow orchestration method, and has the same or similar beneficial effects. To avoid repetition, it will not be described here.
[0080] It should be noted that the workflow orchestration device provided in the embodiment of the present application and the workflow orchestration method provided in the embodiment of the present application are based on the same application concept, so the specific implementation of this embodiment can refer to the implementation of the aforementioned workflow orchestration method, and has the same or similar beneficial effects, and the repeated parts will not be repeated.
[0081] Corresponding to the workflow arrangement method provided in the above embodiment, based on the same technical concept, the embodiment of the present application also provides an electronic device, which is used to execute the above workflow arrangement method. Figure 7 A schematic diagram of the structure of an electronic device for implementing each embodiment of the present application is shown in FIG. Figure 7 As shown. The electronic device may have relatively large differences due to different configurations or performances, and may include one or more processors 701 and memory 702, and the memory 702 may store one or more storage applications or data. Among them, the memory 702 may be a temporary storage or a permanent storage. The application stored in the memory 702 may include one or more modules (not shown in the figure), and each module may include a series of computer executable instructions in the electronic device.
[0082] Furthermore, the processor 701 can be configured to communicate with the memory 702 to execute a series of computer executable instructions in the memory 702 on the electronic device. The electronic device may also include one or more power supplies 703, one or more wired or wireless network interfaces 704, one or more input and output interfaces 705, and one or more keyboards 706.
[0083] Specifically in this embodiment, the electronic device includes a processor, a communication interface, a memory and a communication bus; wherein the processor, the communication interface and the memory communicate with each other through the bus; the memory is used to store computer programs; the processor is used to execute the programs stored in the memory to achieve the above Figure 1 The various steps in the method embodiment are similar to those in the method embodiment, and have the beneficial effects of the above method embodiments. To avoid repetition, the embodiments of the present application will not be described again here.
[0084] This embodiment also provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the above Figure 1 Each step in the method embodiment has the beneficial effects of the above method embodiments. To avoid repetition, the embodiments of the present application will not be described again here.
[0085] The present application provides a computer program product, which includes a computer program stored on a non-transitory computer-readable storage medium. The computer program includes program instructions. When the program instructions are executed by a computer, the computer executes the above Figure 1 Each step in the method embodiment has the beneficial effects of the above method embodiments. To avoid repetition, the embodiments of the present application will not be described again here.
[0086] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, devices, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.
[0087] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.
[0088] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.
[0089] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process in the computer or other programmable device. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.
[0090] In a typical configuration, an electronic device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.
[0091] The memory may include non-permanent storage in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.
[0092] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.
[0093] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.
[0094] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, devices or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment or an embodiment combining software and hardware. Moreover, the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes.
[0095] The above are only embodiments of the present application and are not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included within the scope of the claims of the present application.
Claims
1. A workflow arrangement method, characterized in that: The workflow arrangement method includes: Obtain target voice service for multimedia calls; In response to a first input from a user, selecting a first target capability corresponding to the target voice service from a capability library of an artificial intelligence; Arranging a calling sequence of the first target capability based on the voice service demand of the target voice service to obtain a workflow; The first target capability is called according to the calling sequence in the workflow to complete the target voice service.
2. The workflow arrangement method according to claim 1, characterized in that: Before selecting, in response to the first input of the user, a first target capability corresponding to the target voice service from the capability library of the artificial intelligence, the method further includes: Each capability of the artificial intelligence is registered in the capability library of the artificial intelligence.
3. The workflow arrangement method according to claim 2, characterized in that: The registering the capabilities of the artificial intelligence in the capability library of the artificial intelligence includes: Each of the capabilities is registered in the capability library of the artificial intelligence based on the hypertext transfer protocol.
4. The workflow arrangement method according to claim 2, characterized in that: The registering the capabilities of the artificial intelligence in the capability library of the artificial intelligence includes: The full-duplex communication protocol based on the transmission control protocol registers each of the capabilities in the capability library of the artificial intelligence.
5. The workflow arrangement method according to any one of claims 1 to 4, characterized in that: After arranging the calling sequence of the first target capability based on the voice service demand of the target voice service to obtain a workflow, the method further includes: In response to a second input from the user, selecting a second target capability from the capability library of the artificial intelligence; The first target capability is arranged based on the voice service demand of the target voice service to obtain a workflow including: The calling sequence of the first target capability and the second target capability is arranged based on the voice service demand of the target voice service to obtain the workflow.
6. The workflow arrangement method according to any one of claims 1 to 4, characterized in that: After arranging the calling sequence of the first target capability based on the voice service demand of the target voice service to obtain a workflow, the method further includes: In response to a third input from the user, the order in which the first target capabilities are called in the workflow is adjusted.
7. The workflow arrangement method according to any one of claims 1 to 4, characterized in that: After arranging the calling sequence of the first target capability based on the voice service demand of the target voice service to obtain a workflow, the method further includes: In response to a fourth input from the user, part of the first target capabilities in the workflow is deleted.
8. The workflow arrangement method according to any one of claims 1 to 4, characterized in that: The calling sequence of the first target capability is arranged based on the voice service demand of the target voice service to obtain a workflow including: Determining an arrangement mode of the first target capability based on a voice service demand of the target voice service; Arrange the calling sequence of the first target capability according to the arrangement mode to obtain a workflow, wherein the arrangement mode includes: group arrangement, serial arrangement, parallel arrangement and conditional arrangement; The serial arrangement indicates that the first target capabilities are called sequentially, the parallel arrangement indicates that the first target capabilities are called in parallel, the conditional arrangement indicates that the first target capabilities are called according to preset conditions, and the group arrangement indicates that the first target capabilities are called in groups.
9. The workflow arrangement method according to claim 8, characterized in that: The grouping arrangement includes: The first target capabilities with high correlation are divided into a group, and the first target capabilities in the group are called in parallel.
10. The workflow arrangement method according to claim 1, characterized in that: Before arranging the calling sequence of the first target capability based on the voice service demand of the target voice service to obtain the workflow, the method further includes: Initiate a debugging request to the browser front-end page; Calling the inference platform backend interface through the browser frontend page; The orchestration service is called through the inference platform backend interface to execute a custom script to debug the orchestration environment.
Citation Information
Patent Citations
Intelligent voice AI capability unified scheduling system
CN114187908A
Service capability using method, device and equipment
CN115202623A
Service orchestration method and device, electronic equipment and storage medium
CN115379421A
Order task processing method and device in MES system, medium and equipment
CN116167590A
Atomic service orchestration method and device, equipment and storage medium
CN116755677A
