Dynamic workflow generation method and system

Abstract

The invention discloses a dynamic workflow generation method and system, and the method comprises the steps: obtaining a configuration parameter of a target workflow, the workflow being a workflow in a generative artificial intelligence application; using the configuration parameters to configure parameters in a workflow; and executing the workflow of the configuration parameters to generate an artificial intelligence object. According to the dynamic workflow generation method and system, for the generative artificial intelligence application workflow, if the generation target is adjusted, the generation target in the workflow can be modified into a new style by configuring the workflow parameter JSON data in the background, a workflow calling program does not need to be modified, and the working efficiency is improved.

Description

Technical Field

[0001] The invention relates to a dynamic workflow generation method and system. Background Art

[0002] Generative AI is a type of technology based on deep learning that can autonomously generate content such as text, images, audio or code. Its core lies in training models through massive data to learn data distribution and simulate human creativity to generate new samples.

[0003] Generative AI has been applied in many fields:

[0004] Content creation: Automatically generate marketing copy, design sketches and video materials;

[0005] Industrial design: Generate 3D models of parts based on parameter constraints;

[0006] Medical assistance: synthesizing medical imaging data to enhance diagnostic model training;

[0007] Code generation: Completing programming code in real time based on context.

[0008] In existing generative artificial intelligence, workflows generally have pre-defined fixed calling parameters. Once new requirements arise, not only the internal logic of the workflow needs to be adjusted, but the calling parameters also need to be modified, which makes maintenance and expansion difficult. Summary of the Invention

[0009] The technical problem to be solved by the present invention is to overcome the defect that workflows in the prior art generally have fixed calling parameters pre-defined. Once new requirements arise, not only the internal logic of the workflow needs to be adjusted, but also the calling parameters need to be modified, which is difficult to maintain and expand. A dynamic workflow generation method and system are provided, which can improve work efficiency by configuring workflow parameter JSON data in the background and modifying the generated objects therein to a new style without modifying the workflow calling program.

[0010] This application solves the above technical problems through the following technical solutions:

[0011] A dynamic workflow generation method, characterized in that the dynamic workflow generation method comprises:

[0012] Obtaining configuration parameters of a target workflow, where the workflow is a workflow in a generative artificial intelligence application;

[0013] Configuring parameters in the workflow using the configuration parameters;

[0014] Execute the workflow of configuration parameters to generate artificial intelligence objects.

[0015] Preferably, the step of obtaining the configuration parameters of the target workflow includes:

[0016] Get the configuration target of the artificial intelligence object;

[0017] Generate configuration parameters for the configuration target using JSON format, the configuration parameters including workflow node data and trigger conditions;

[0018] Store configuration parameters in the server;

[0019] The configuration parameters of the target workflow are obtained from the server.

[0020] Preferably, configuring parameters in the workflow using the configuration parameters includes:

[0021] Parse the configuration parameters in JSON format to obtain data fields;

[0022] Find the configuration address corresponding to the data field;

[0023] Use the configuration address to configure the parameters in the workflow.

[0024] Preferably, the step of obtaining the configuration parameters of the target workflow includes:

[0025] Visualize the workflow to obtain the modular flow chart corresponding to the workflow;

[0026] The modular flow chart is used to generate configuration parameters for the configuration target, wherein the configuration parameters are imperative configuration parameters in JSON format.

[0027] Preferably, the dynamic workflow generation method includes:

[0028] Get the target workflow process;

[0029] Get the configuration node in the target workflow;

[0030] Dividing the target workflow into a plurality of workflow modules according to the configuration nodes, each workflow module including at least one configuration node;

[0031] The divided workflow modules are visualized to generate a modular flow chart. Each graphic module in the modular flow chart includes at least one input node, and the input node corresponds to the configuration node in a one-to-one manner.

[0032] Preferably, the generating of configuration parameters for the configuration target by using the modularized flowchart includes:

[0033] Collect configuration parameters of each configuration node;

[0034] Use the configuration node to generate the configuration address in the configuration parameters;

[0035] Generates the configuration parameters with the configuration address into imperative configuration parameters in JSON format.

[0036] Preferably, the dynamic workflow generation method includes:

[0037] Determine whether the workflow corresponding to the modular flowchart has changed, and if so, obtain the difference between the graphic modules;

[0038] Determine whether the configuration parameters before the change are still operational. If not, use artificial intelligence to generate compensation parameters for the difference.

[0039] The compensation parameters are used to generate a changed workflow and a modular flow chart of the changed workflow.

[0040] A dynamic workflow generation system is characterized in that the dynamic workflow generation system is used to implement the dynamic workflow generation method described above.

[0041] On the basis of conforming to the common sense in this field, the above-mentioned preferred conditions can be arbitrarily combined to obtain the preferred embodiments of the present invention.

[0042] The positive progress effect of the present invention is:

[0043] The dynamic workflow generation method and system of the present invention are aimed at generative artificial intelligence application workflows. If there is an adjustment to the generation target, the generated object can be modified to a new style by configuring the workflow parameter JSON data in the background, without modifying the workflow calling program, thereby improving work efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 This is a flow chart of a dynamic workflow generation method according to embodiment 1 of the present invention. DETAILED DESCRIPTION

[0045] The present invention is further described below by way of examples, but the present invention is not limited to the scope of the examples.

[0046] Example 1

[0047] This embodiment provides a dynamic workflow generation system, which includes a server.

[0048] The server can be an independent physical server, or a server cluster or distributed system composed of multiple physical servers. It can also be a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, CDN (Content Delivery Network), as well as big data and artificial intelligence platforms.

[0049] The server is used to:

[0050] Obtaining configuration parameters of a target workflow, where the workflow is a workflow in a generative artificial intelligence application;

[0051] Configuring parameters in the workflow using the configuration parameters;

[0052] Execute the workflow of configuration parameters to generate artificial intelligence objects.

[0053] Specifically, the server is used to:

[0054] Get the configuration target of the artificial intelligence object;

[0055] Generate configuration parameters for the configuration target using JSON format, the configuration parameters including workflow node data and trigger conditions;

[0056] Store configuration parameters in the server;

[0057] The configuration parameters of the target workflow are obtained from the server.

[0058] The configuration parameters of the workflow are predefined as a JSON string (which may include multiple parameter fields). The JSON data is configured in the service background. The workflow initialization module reads the pre-configured JSON data passed in.

[0059] Parse the corresponding fields and execute the corresponding dynamic workflow logic as needed.

[0060] Specifically, the server is used to:

[0061] Parse the configuration parameters in JSON format to obtain data fields;

[0062] Find the configuration address corresponding to the data field;

[0063] Use the configuration address to configure the parameters in the workflow.

[0064] Specifically, the server is used to:

[0065] Visualize the workflow to obtain the modular flow chart corresponding to the workflow;

[0066] The modular flow chart is used to generate configuration parameters for the configuration target, wherein the configuration parameters are imperative configuration parameters in JSON format.

[0067] The current encoding method is declarative encoding, which is inefficient and may lead to inaccurate retrieval judgment of descriptive words.

[0068] After the workflow is visualized in this embodiment, it is convenient for users to edit and use the edited editing position to quickly generate configuration parameters corresponding to the workflow configuration address, further improving the efficiency and accuracy of the workflow.

[0069] Specifically, the server is used to:

[0070] Get the target workflow process;

[0071] Get the configuration node in the target workflow;

[0072] Dividing the target workflow into a plurality of workflow modules according to the configuration nodes, each workflow module including at least one configuration node;

[0073] The divided workflow modules are visualized to generate a modular flow chart. Each graphic module in the modular flow chart includes at least one input node, and the input node corresponds to the configuration node in a one-to-one manner.

[0074] This embodiment modularizes the workflow to facilitate the selection and operation of configuration nodes. At the same time, after the node is selected, it is easy to generate configuration parameters, further improving efficiency.

[0075] Wherein, the server is used for:

[0076] Collect configuration parameters of each configuration node through the input terminal;

[0077] Use the configuration node to generate the configuration address in the configuration parameters;

[0078] Generates the configuration parameters with the configuration address into imperative configuration parameters in JSON format.

[0079] Wherein, the server is used for:

[0080] Determine whether the workflow corresponding to the modular flowchart has changed, and if so, obtain the difference between the graphic modules;

[0081] Determine whether the configuration parameters before the change are still operational. If not, use artificial intelligence to generate compensation parameters for the difference.

[0082] If the workflow is changed or upgraded, the original configuration address will become invalid. This embodiment uses a modular flowchart to facilitate finding change differences, and the differences between modules can be transitioned and compensated using artificial intelligence. The use of the transition module can adapt the configuration parameters that have not been upgraded, thereby improving production efficiency.

[0083] The compensation parameters are used to generate a changed workflow and a modular flow chart of the changed workflow.

[0084] See also Figure 1 , using the above-mentioned dynamic workflow generation system, this embodiment also provides a dynamic workflow generation method, including:

[0085] Step 100: Obtain configuration parameters of a target workflow, where the workflow is a workflow in a generative artificial intelligence application;

[0086] Step 101: Use the configuration parameters to configure the parameters in the workflow;

[0087] Step 102: Execute the workflow of configuration parameters to generate artificial intelligence objects.

[0088] Specifically, the dynamic workflow generation method further includes:

[0089] Step 200: Obtain the configuration target of the artificial intelligence object;

[0090] Step 201: Generate configuration parameters for the configuration target using JSON format, where the configuration parameters include node data and trigger conditions of the workflow;

[0091] Step 202: store the configuration parameters in the server;

[0092] Step 100: Obtain configuration parameters of the target workflow from the server.

[0093] Step 101 specifically includes:

[0094] Step 1011: Parse the configuration parameters in JSON format to obtain data fields;

[0095] Step 1012: Find the configuration address corresponding to the data field;

[0096] Step 1013: Use the configuration address to configure the parameters in the workflow.

[0097] Step 100 specifically includes:

[0098] Visualize the workflow to obtain the modular flow chart corresponding to the workflow;

[0099] The modular flow chart is used to generate configuration parameters for the configuration target, wherein the configuration parameters are imperative configuration parameters in JSON format.

[0100] The step of visualizing the workflow to obtain a modular flowchart corresponding to the workflow includes:

[0101] Get the target workflow process;

[0102] Get the configuration node in the target workflow;

[0103] Dividing the target workflow into a plurality of workflow modules according to the configuration nodes, each workflow module including at least one configuration node;

[0104] The divided workflow modules are visualized to generate a modular flow chart. Each graphic module in the modular flow chart includes at least one input node, and the input node corresponds to the configuration node in a one-to-one manner.

[0105] Furthermore, the generating of configuration parameters for the configuration target by using the modularized flowchart includes:

[0106] Collect configuration parameters of each configuration node;

[0107] Use the configuration node to generate the configuration address in the configuration parameters;

[0108] Generates the configuration parameters with the configuration address into imperative configuration parameters in JSON format.

[0109] The dynamic workflow generation method includes:

[0110] Determine whether the workflow corresponding to the modular flowchart has changed, and if so, obtain the difference between the graphic modules;

[0111] Determine whether the configuration parameters before the change are still operational. If not, use artificial intelligence to generate compensation parameters for the difference.

[0112] The compensation parameters are used to generate a changed workflow and a modular flow chart of the changed workflow.

[0113] Although specific embodiments of the present invention have been described above, those skilled in the art will appreciate that these are merely illustrative and that the scope of the present invention is defined by the appended claims. Those skilled in the art may make various changes or modifications to these embodiments without departing from the principles and essence of the present invention, and such changes and modifications are intended to fall within the scope of the present invention.

Claims

1. A dynamic workflow generation method, characterized in that: The dynamic workflow generation method comprises: Obtaining configuration parameters of a target workflow, where the workflow is a workflow in a generative artificial intelligence application; Configuring parameters in the workflow using the configuration parameters; Execute the workflow of configuration parameters to generate artificial intelligence objects.

2. The dynamic workflow generation method according to claim 1, wherein: The step of obtaining the configuration parameters of the target workflow includes: Get the configuration target of the artificial intelligence object; Generate configuration parameters for the configuration target using JSON format, the configuration parameters including workflow node data and trigger conditions; Store configuration parameters in the server; The configuration parameters of the target workflow are obtained from the server.

3. The dynamic workflow generation method according to claim 2, wherein: The configuring parameters in the workflow using the configuration parameters includes: Parse the configuration parameters in JSON format to obtain data fields; Find the configuration address corresponding to the data field; Use the configuration address to configure the parameters in the workflow.

4. The dynamic workflow generation method according to claim 2, wherein: The step of obtaining the configuration parameters of the target workflow includes: Visualize the workflow to obtain the modular flow chart corresponding to the workflow; The modular flow chart is used to generate configuration parameters for the configuration target, wherein the configuration parameters are imperative configuration parameters in JSON format.

5. The dynamic workflow generation method according to claim 4, wherein: The dynamic workflow generation method comprises: Get the target workflow process; Get the configuration node in the target workflow; Dividing the target workflow into a plurality of workflow modules according to the configuration nodes, each workflow module including at least one configuration node; The divided workflow modules are visualized to generate a modular flow chart. Each graphic module in the modular flow chart includes at least one input node, and the input node corresponds to the configuration node in a one-to-one manner.

6. The dynamic workflow generation method according to claim 5, wherein: The generating of configuration parameters for the configuration target by using the modularized flowchart includes: Collect configuration parameters of each configuration node; Use the configuration node to generate the configuration address in the configuration parameters; Generates the configuration parameters with the configuration address into imperative configuration parameters in JSON format.

7. The dynamic workflow generation method according to claim 5, wherein: The dynamic workflow generation method comprises: Determine whether the workflow corresponding to the modular flowchart has changed, and if so, obtain the difference between the graphic modules; Determine whether the configuration parameters before the change are still operational. If not, use artificial intelligence to generate compensation parameters for the difference. The compensation parameters are used to generate a changed workflow and a modular flow chart of the changed workflow.

8. A dynamic workflow generation system, characterized in that: The dynamic workflow generation system is used to implement the dynamic workflow generation method according to any one of claims 1 to 7.

Images