An instant delivery AI customer service evaluation use case generation method, device, equipment and medium

Abstract

The application discloses an instant delivery AI customer service evaluation case generation method and device, equipment and medium. The method comprises the following steps: performing request analysis on an obtained evaluation case generation request, determining target state information, a test environment identifier and associated parameter information; determining a target test environment according to the test environment identifier, calling a delivery personnel registration interface of the target test environment to construct a delivery personnel entity, and determining a delivery personnel identifier corresponding to the delivery personnel entity; calling an order testing case of an interface testing platform to execute in the target test environment, obtaining a test result, and determining an order identifier according to the test result; generating an evaluation case according to the target state information, the delivery personnel identifier, the order identifier and the associated parameter information; the evaluation case is used for being imported into an artificial intelligence (AI) customer service evaluation platform running in the target test environment, and an AI customer service intelligent agent of a generative AI application development platform is driven to execute instant delivery AI customer service simulation evaluation by the AI customer service evaluation platform.

Description

Technical Field

[0001] This invention relates to the field of testing technology, and in particular to a method, apparatus, equipment, and medium for generating evaluation test cases for AI customer service in real-time delivery. Background Technology

[0002] In evaluating AI (Artificial Intelligence) customer service for on-demand delivery, it is necessary to prepare real orders in specific business states for each test case, such as "order accepted," "sender contacted," "on the spot," and "completed." Traditionally, testers manually register or select test riders and construct orders, then gradually modify the order status using internal tools before backfilling key fields into the evaluation test cases.

[0003] Traditional AI customer service test cases for instant delivery involve a lengthy manual data generation process, resulting in low efficiency in test case generation. Furthermore, different order states require different state paths, necessitating testers to manually memorize the links. When memory bias exists, it reduces the accuracy of test case generation, leading to low accuracy in the final test results for AI customer service in instant delivery. Summary of the Invention

[0004] This invention provides a method, apparatus, device, and medium for generating test cases for AI customer service evaluation in instant delivery, so as to improve the efficiency and accuracy of test case generation in the scenario of AI customer service evaluation for instant delivery.

[0005] According to one aspect of the present invention, a method for generating evaluation test cases for AI customer service in instant delivery is provided, the method comprising: In response to the tester's request to generate test cases, the request is parsed to determine the target status information, test environment identifier, and associated parameter information; The target test environment is determined based on the test environment identifier. The delivery personnel registration interface of the target test environment is called to construct the delivery personnel entity and the delivery personnel identifier corresponding to the delivery personnel entity is determined. The order placement test cases of the API testing platform are executed in the target testing environment to obtain test results, and the order identifier is determined based on the test results. Based on the target status information, the delivery personnel identifier, the order identifier, and the associated parameter information, evaluation cases are generated. The evaluation cases are imported into the AI ​​customer service evaluation platform running in the target test environment, and the AI ​​customer service agent of the generative AI application development platform is driven by the AI ​​customer service evaluation platform to perform real-time delivery AI customer service simulation evaluation.

[0006] According to another aspect of the present invention, an apparatus for generating evaluation test cases for AI-powered instant delivery customer service is provided, the apparatus comprising: The request and response module is used to respond to the testers' request to generate test cases, parse the request to determine the target status information, test environment identifier and associated parameter information; The personnel identification module is used to determine the target test environment based on the test environment identifier, call the delivery personnel registration interface of the target test environment to construct a delivery personnel entity, and determine the delivery personnel identifier corresponding to the delivery personnel entity; The order identifier determination module is used to call the order placement test cases of the interface test platform to execute in the target test environment, obtain the test results, and determine the order identifier based on the test results; The evaluation test case generation module is used to generate evaluation test cases based on the target status information, the delivery personnel identifier, the order identifier, and the associated parameter information. The evaluation test cases are imported into the AI ​​customer service evaluation platform running in the target test environment, and the AI ​​customer service intelligent agent of the generative AI application development platform is driven by the AI ​​customer service evaluation platform to perform real-time delivery AI customer service simulation evaluation.

[0007] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising: At least one processor; and A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to execute the instant delivery AI customer service evaluation test case generation method according to any embodiment of the present invention.

[0008] According to another aspect of the present invention, a computer-readable storage medium is provided, the computer-readable storage medium storing computer instructions for causing a processor to execute and implement the instant delivery AI customer service evaluation test case generation method according to any embodiment of the present invention.

[0009] The technical solution of this invention determines the target test environment based on the test environment identifier, constructs a delivery person entity by calling the delivery person registration interface of the target test environment, and determines the delivery person identifier corresponding to the delivery person entity; executes the order placement test case of the interface test platform in the target test environment to obtain the test result, and determines the order identifier based on the test result; and generates evaluation test cases based on the target status information, delivery person identifier, order identifier, and associated parameter information. The above technical solution realizes the automated generation of test cases in the instant delivery AI customer service evaluation scenario. By automatically generating data based on the status identifier given by the tester, it advances the order execution status, improves the efficiency of test case generation in the instant delivery AI customer service evaluation scenario, and improves the accuracy of test case generation in the instant delivery AI customer service evaluation scenario by eliminating the need for manual memorization of the chain.

[0010] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of the present invention, nor is it intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description

[0011] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0012] Figure 1 This is a flowchart of a method for generating evaluation test cases for AI-powered instant delivery customer service, according to Embodiment 1 of the present invention. Figure 2A This is a flowchart of a method for generating evaluation test cases for AI-powered instant delivery customer service, according to Embodiment 2 of the present invention. Figure 2B This is a schematic diagram of the structure of a directed graph of order status provided in Embodiment 2 of the present invention; Figure 3 This is a flowchart of a method for generating evaluation test cases for AI-powered instant delivery customer service, according to Embodiment 3 of the present invention. Figure 4 This is a schematic diagram of the structure of an instant delivery AI customer service evaluation test case generation device according to Embodiment 4 of the present invention; Figure 5 This is a schematic diagram of the structure of an electronic device that implements the instant delivery AI customer service evaluation test case generation method of this invention. Detailed Implementation

[0013] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0014] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0015] Example 1 Figure 1 This is a flowchart of a method for generating test cases for AI customer service evaluation in instant delivery, provided in Embodiment 1 of the present invention. This embodiment is applicable to the automated generation of test cases for AI customer service evaluation scenarios with multi-turn dialogues. This method can be executed by an instant delivery AI customer service test case generation device, which can be implemented in hardware and / or software and can be configured in an electronic device. Figure 1 As shown, the method includes: S110. In response to the tester's request to generate test cases, the request is parsed to determine the target status information, test environment identifier, and associated parameter information.

[0016] S120. Determine the target test environment based on the test environment identifier, call the delivery personnel registration interface of the target test environment to construct the delivery personnel entity, and determine the delivery personnel identifier corresponding to the delivery personnel entity.

[0017] S130. The order placement test cases of the call interface test platform are executed in the target test environment to obtain test results, and the order identifier is determined based on the test results.

[0018] S140. Based on the target status information, delivery personnel identification, order identification, and associated parameter information, generate evaluation cases. The evaluation cases are imported into the AI ​​customer service evaluation platform running in the target test environment. The AI ​​customer service evaluation platform drives the AI ​​customer service agent of the generative AI application development platform to perform real-time delivery AI customer service simulation evaluation.

[0019] When testers need to evaluate AI-powered customer service for instant delivery, they can initiate a test case generation request. When initiating this request, testers include target status information, a test environment identifier, and associated parameter information. The target status information determines the current order status during the test, such as order accepted, delivery personnel on-site, recipient contacted, and order completion. The test environment identifier uniquely identifies the test environment; the associated parameter information serves as the input parameters for the interfaces called during the subsequent construction of the delivery personnel entity and order generation process. The target status information, test environment identifier, and associated parameter information are determined through request parsing of the test case generation request.

[0020] For example, let's illustrate the target status information using a specific scenario. Consider a tester wanting to test a multi-round AI customer service conversation scenario for instant delivery where "the user initiates an address change after the package has been picked up." In this scenario, the target status information could be the status code corresponding to "package picked up," or it could be the complete order execution chain given by the tester, from the initial order state to the "package picked up" state. If the target status information only contains the status code corresponding to "package picked up," the complete order execution chain can be automatically generated based on the target status information.

[0021] Based on the test environment identifier, the target test environment required for execution can be determined, and rider entities and test orders can be constructed within the target test environment. Specifically, the delivery personnel registration interface of the target test environment is called to construct the delivery personnel entity. The input parameters of the delivery personnel registration interface can include delivery personnel identity information, city / region identifier, personnel training location identifier, order type, test environment identifier, and initial delivery personnel status. It should be noted that the required input parameters when calling the relevant interfaces can be obtained from the associated parameter information. The initial delivery personnel status is the default normal status; other statuses include frozen, expired, banned, permanently banned, and forced termination.

[0022] When the delivery personnel registration interface returns a successful registration result, the delivery personnel identifier (courier_id) corresponding to the successfully registered delivery personnel entity can be retrieved using the registered personnel's identity information. This identifier is used to characterize the uniqueness of the delivery personnel. The order placement test cases of the interface testing platform are executed within the target test environment to obtain the test results, and the order identifier is determined based on the test results.

[0023] Specifically, the order placement test cases on the interface testing platform can be, for example, the Open Platform V5 order placement test cases. Calling the order placement test case switches to the target testing environment, passing in relevant order test parameters to the test case interface. These parameters can include, for example, merchant information, current timestamp, city identifier, sender information, recipient information, order type, appointment time, merchant store identifier, travel method, whether a delivery code is required, item information, item type, weight, item value, insurance information, and order address. Calling the test case executes the interface to trigger asynchronous order placement and obtains the test results.

[0024] After successful test case execution, the returned JSON (JavaScript Object Notation) format "response_body" is parsed, and the order identifier "orderNumber" is extracted from "response_body". This identifier is used to characterize the uniqueness of the order. In other words, a brand new blank test order is generated in the test environment, with an initial state of 0 and no state transitions.

[0025] Based on the target status information, delivery personnel identifier, order identifier, and associated parameter information, test cases are generated. Specifically, based on the target status information, the target order status of the order to be tested can be determined, such as "picked up" or "completed." Order path planning is then performed based on the target order status. For example, if the target order status is "picked up," the planned path for the instant delivery order scenario is "order accepted → sender contacted → delivery personnel on standby → picked up." The state progression is then executed according to the planned path within the target test environment. Within the order center of the target test environment, a real order with the `orderNumber` exists, and its status within the environment is "picked up," while also being bound to the delivery personnel identifier `courier_id`.

[0026] The evaluation test cases are generated by filling the evaluation test case template with the delivery personnel identifier (courier_id), order identifier (orderNumber), order latitude and longitude information, delivery personnel latitude and longitude information, and city identifier from the associated parameter information as agent_inputs. During the actual testing phase, the evaluation test cases are imported into the AI ​​customer service evaluation platform running in the target testing environment. The AI ​​customer service evaluation platform initiates a session request, passing the agent_inputs from the evaluation test cases to the AI ​​customer service agent of the generative AI application development platform, starting a simulated multi-round session. The AI ​​customer service evaluation platform can collect the evaluation reports output from each round to further determine the evaluation results of the instant delivery AI customer service.

[0027] The technical solution of this invention determines the target test environment based on the test environment identifier, constructs a delivery person entity by calling the delivery person registration interface of the target test environment, and determines the delivery person identifier corresponding to the delivery person entity; executes the order placement test case of the interface test platform in the target test environment to obtain the test result, and determines the order identifier based on the test result; and generates evaluation test cases based on the target status information, delivery person identifier, order identifier, and associated parameter information. The above technical solution realizes the automated generation of test cases in the instant delivery AI customer service evaluation scenario. By automatically generating data based on the status identifier given by the tester, it advances the order execution status, improves the efficiency of test case generation in the instant delivery AI customer service evaluation scenario, and improves the accuracy of test case generation in the instant delivery AI customer service evaluation scenario by eliminating the need for manual memorization of the chain.

[0028] Example 2 Figure 2A This is a flowchart of a method for generating evaluation test cases for AI customer service in instant delivery, provided in Embodiment 2 of the present invention. This embodiment is an optimization and improvement based on the above technical solutions.

[0029] Furthermore, the step "generate test cases based on target status information, delivery personnel identifier, order identifier, and associated parameter information" is refined to "if the target status information is a target status code, then based on the target status code and a pre-built directed graph of order status, determine the target status path; call the status advancement tool of the target test environment, and based on the target status path, order identifier, and delivery personnel identifier, advance the order status within the target test environment to ensure that the order status within the target test environment is the order status corresponding to the target status code; generate test cases based on the order identifier, delivery personnel identifier, and associated parameter information." This improves the method for generating test cases.

[0030] It should be noted that for parts not described in detail in this embodiment of the invention, please refer to the descriptions in other embodiments. To further achieve automated data generation and avoid the long processes, numerous operations, and easy omissions associated with manual data generation, this embodiment provides an automated method for generating target state paths to replace manual methods, such as... Figure 2A As shown, the method includes the following specific steps: S210. In response to the tester's request to generate test cases, the request is parsed to determine the target status information, test environment identifier, and associated parameter information.

[0031] S220. Determine the target test environment based on the test environment identifier, call the delivery personnel registration interface of the target test environment to construct the delivery personnel entity, and determine the delivery personnel identifier corresponding to the delivery personnel entity.

[0032] S230. The order placement test cases of the call interface test platform are executed in the target test environment to obtain test results, and the order identifier is determined based on the test results.

[0033] S240. If the target status information is a target status code, then the target status path is determined based on the target status code and the pre-built directed graph of order status.

[0034] S250: Invoke the status advancement tool of the target test environment. Based on the target status path, order identifier, and delivery personnel identifier, advance the order status within the target test environment to ensure that the order status within the target test environment is the order status corresponding to the target status code.

[0035] S260. Based on the order identifier, delivery personnel identifier, and associated parameter information, generate test cases. The test cases are imported into the AI ​​customer service evaluation platform running in the target test environment. The AI ​​customer service evaluation platform drives the AI ​​customer service agent of the generative AI application development platform to perform real-time delivery AI customer service simulation evaluation.

[0036] If the target status information provided by the tester is a target status code, which only indicates the target order status within the target test environment as desired by the tester, then a target status path can be automatically generated based on the target status code. The target status path is generated based on a pre-built directed graph of order status. Different target states require different preceding entities and status paths; relying on the tester's memory of the links is unmaintainable. Therefore, this embodiment also provides a method for generating a directed graph of order status. In an optional embodiment, the directed graph of order status is constructed as follows: Determine several order execution states in an instant order delivery scenario and construct a status code corresponding to each order execution state; identify the status code corresponding to each order execution state as a state node; construct state node edges between each state node based on the association between each order execution state; generate metadata for the state node edges; the metadata includes the starting status code, target status code, transition preconditions, transition dependency parameter fields, idempotent keys, and transition failure compensation operation information; construct a directed graph of order states including each state node and its corresponding state node edges.

[0037] Specifically, order execution status includes: Order accepted, Sender contacted, Delivery personnel on standby, Package picked up, Recipient contacted, Completed, and Cancelled. To further simplify the directed graph of order statuses and ensure the efficiency of subsequent graph search processes, different status codes can be used to represent different order execution statuses. Status codes corresponding to different order execution statuses are constructed. For example, the status code for "Order accepted" is set to "30", the status code for "Sender contacted" is set to "40", the status code for "Delivery personnel on standby" is set to "42", the status code for "Packed" is set to "44", the status code for "Recipient contacted" is set to "52", the status code for "Completed" is set to "60", and the status code for "Cancelled (order accepted and then cancelled)" is set to "64".

[0038] The status codes corresponding to each order's execution status are used as state nodes in a directed graph. These state nodes have directed relationships related to the order execution process. Specifically, the complete order execution chain is: "Initial Order State → Order Accepted → Sender Called → Delivery Personnel Ready → Package Picked Up → Recipient Called → Completed" and "Initial Order State → Order Accepted → Sender Called → Delivery Personnel Ready → Cancelled." See also: Figure 2B The diagram shows a structure of a directed graph of order status.

[0039] Based on the order execution statuses with existing relationships, state node edges are constructed between nodes. The metadata for each state node edge includes the starting status code (from), the target status code (to), the transition preconditions (condition), the transition dependency parameter field (requires), the idempotency key (idempotency_key), and the transition failure compensation operation information (compensate). The idempotency key (idempotency_key) and the transition failure compensation operation information (compensate) are set to null and marked as expandable fields when the current tool lacks the corresponding capabilities.

[0040] In this context, the starting status code (from) is the starting node corresponding to the status node edge; the target status code (to) is the target node corresponding to the status node edge. For example, the starting status code (from) of the status node edge between node "42" and node "44" is "42", and the target status code (to) is "44". The transition precondition (condition) is the condition that needs to be completed when the order execution state corresponding to the target status code is reached. For example, the transition precondition (condition) of the status node edge between node "42" and node "44" can be "Pickup code verification passed"; another example is that the transition precondition (condition) of the status node edge between node "30" and node "40" can be "Sender's phone number is valid", etc. This embodiment will not elaborate further on these.

[0041] The transition dependency parameter field can be any parameter field that determines the transition from one node state to the next during the order state progression. For example, the `requires` parameter field for the transition dependency parameter edge between node "42" and node "44" could include "courier_id"; the `requires` parameter field for the transition dependency parameter edge between node "30" and node "40" could include "sender_phone". In other words, the parameter fields required for transitions between different node states are different.

[0042] The idempotency key is used to prevent duplicate order calls. When constructing the directed graph of order states, the idempotency key is only a placeholder field; it is not an actual usable key value. It is a rule template recording the idempotency key, such as "MD5({orderNumber}_{to_status_code})". This template is dynamically generated in subsequent state progression phases. Different idempotency keys need to be generated for the same transition edge in different orders.

[0043] The compensation operation information for transition failures (compensate) refers to the compensation operations performed after a transition failure, such as reverting to the previous state or clearing temporary data. During the generation phase of the directed graph of order status, the compensation operation information for transition failures only defines the configuration or template of the compensation action. In the subsequent chained state advancement phase, when a transition failure occurs, the compensation logic of the compensation operation can be triggered.

[0044] Construct a directed graph of order states, including each state node and its corresponding state node edges, and detect whether there are isolated states, such as non-final states without incoming or outgoing edges, and check whether there are loops. If there are any construction anomalies, an alarm will be output so that testers can manually intervene and make changes.

[0045] If the target state information provided by the tester is a target state code, then the target state path is obtained by traversing the directed graph of the order status using the Breadth-First Search (BFS) shortest path algorithm based on the target state code. For example, if the target state code is "44", then the target state path is "initial state → 30 → 40 → 42 → 44"; and if the target state code is "60", then the target state path is "initial state → 30 → 40 → 42 → 44 → 52 → 60".

[0046] In an optional embodiment, if the target state information is the initial state path, the initial state path is verified according to the directed graph of the order state; if the verification passes, the initial state path is determined as the target state path.

[0047] In this embodiment scenario, the tester specifies an initial state path chain-codes, which is a manually specified sequence of state codes. When the target state information is the initial state path chain-codes manually given by the tester, the manually specified state code sequence is prone to illegal transitions. For example, after an order has been accepted, it cannot directly jump to calling the recipient. Illegal paths may cause data generation failure or generate data that does not conform to the business process. Therefore, it is necessary to verify the accuracy or legality of the given initial state path chain-codes.

[0048] Specifically, starting from the default initial state, the chain-codes of the initial state path given by the tester are checked item by item. When traversing to the current state node, it is checked whether there is an illegal transition to the next state node connected to the current state node. For example, state node "30" cannot directly transition to "52". In addition, the preconditions of the nodes between the current state node and the nodes connected to it need to be verified. If the parameters of the preconditions are missing, the condition is marked as insufficient and prompts for completion. If the initial state path given by the tester passes the verification, the initial state path is determined as the target state path.

[0049] In an optional embodiment, if the verification fails, the initial state path is corrected based on the directed graph of the order status to obtain the corrected state path and error message; the corrected state path and error message are fed back to the tester so that the tester can confirm the result based on the error message and the corrected state path and provide feedback on the confirmation result; if the confirmation result is a path confirmation result for the corrected state path, the corrected state path is determined as the target state path.

[0050] If the initial state path provided by the tester fails validation, it is automatically corrected based on the directed graph of order states. For example, if the initial state path gives a direct jump from state node "30" to "52", a graph traversal of the directed graph shows that the next valid node after state node "30" should be "40". Therefore, the path is corrected, and the corrected state path and error message are output. For example, the error message could be "State node 30 illegally jumps to 52". The corrected state path and error message are then fed back to the tester, who confirms the corrected state path based on the error message. If the corrected state path is confirmed, the confirmation result is fed back.

[0051] Furthermore, if the tester fails to confirm the corrected state path, they can initiate another state path generated by themselves. This initiated state path will still need to be verified until it passes. If a confirmation result is received from the tester, and this confirmation result confirms the corrected state path, then the corrected state path will be designated as the target state path.

[0052] After determining the target status path, the status advancement tool for the target test environment is invoked. Based on the target status path, order identifier, and delivery personnel identifier, the order status is advanced within the target test environment to ensure that the order status within the target test environment corresponds to the order status of the target status code.

[0053] Specifically, the system verifies that all dependent parameters required for the transition preconditions of the node edges corresponding to each state node in the target state path exist. If any are missing, an error message is displayed or the parameters are automatically completed. After verification, state transitions are executed sequentially according to the state code sequence of the target state path. If the current step is successful, the next step is executed, including retrieving the current target state code and generating idempotent keys based on the transition metadata.

[0054] It's important to note that during order status progression, the target test environment's status progression tool or order status update interface needs to be called. Timeouts, jitter, and request retries may occur during these calls. Without idempotent keys, the interface might receive two requests when network issues arise, potentially leading to duplicate state changes and exceptions. With idempotent keys, repeated requests will not execute the same business logic if the interface recognizes the same idempotent key. For example, if the first request succeeds, the second request will simply return success without modifying the order data. Therefore, idempotent keys effectively prevent network retries during chained state progression from causing the same state transition to be executed repeatedly, ensuring that each state transition takes effect at most once, guaranteeing the reliability of the chained progression, and preventing timeout retries at any step from corrupting the order status.

[0055] Furthermore, the next step involves calling the state advancement tool or order status update interface, passing in the target status code, order identifier, idempotent key, and delivery personnel identifier for the target status path. If the interface returns a successful response, this step is recorded as successful, and the next step continues. If the interface returns a failure, the reason for the failure is recorded, the compensation operation given in the transition failure compensation operation information (compensate) is executed, and subsequent order status advancement is terminated. After all steps are completed, the order query interface is called to obtain the real-time order status and compare it with the target status code. If they do not match, the status is marked as inconsistent, and an alarm is triggered. If they match, it can be determined that the current or real-time order status in the target test environment has reached the order status corresponding to the target status code. Each step carries an idempotent key call to the interface to avoid repeated status changes due to network retries, thereby ensuring the consistency of the order advancement operation.

[0056] It's important to note that in scenarios involving multi-turn conversational AI customer service evaluation, the requirement for data cleanliness is extremely high. For example, if the same delivery person handles multiple orders, or if the delivery person's location coordinates do not match the order's shipping address, the AI ​​workflow may remain at the pre-verification stage, leading to distorted evaluation results. Therefore, before generating evaluation test cases, it's necessary to validate the business data within the target test environment to ensure the accuracy of evaluations in multi-turn conversational AI customer service scenarios.

[0057] In an optional embodiment, before generating test cases based on the order identifier, delivery personnel identifier, and associated parameter information, the method further includes: calling the order information acquisition interface of the generative AI application development platform, performing data verification within the target test environment based on the delivery personnel identifier and the order identifier, and obtaining the data verification result; if the data verification result is successful, then generating test cases based on the order identifier, delivery personnel identifier, and associated parameter information.

[0058] The generative AI application development platform can be the Dify platform. Specifically, the Dify platform is a low-code, open-source visual platform capable of building large-scale intelligent agents, chatbots, and business workflows. Developers do not need to write low-level code for large-scale model calls, multi-turn conversations, and tool calls from scratch; they can easily integrate large models with internal enterprise business systems through drag-and-drop configuration. Specifically, the order information retrieval interface of the Dify platform can be getOrderInfo().

[0059] Specifically, by calling the order information retrieval interface getOrderInfo() and passing in the delivery person's identifier and order identifier, all orders and location data under the delivery person's name can be obtained. Based on the data returned by the interface, data verification is performed in the target test environment to obtain the data verification results.

[0060] Optionally, the order information acquisition interface of the generative AI application development platform is called to perform data verification within the target test environment based on the delivery personnel identifier and the order identifier, and obtain the data verification result. This includes: calling the order information acquisition interface of the generative AI application development platform and using the order identifier and the delivery personnel identifier as interface parameters to obtain the current order status, delivery personnel order information, and delivery personnel real-time location information within the target test environment; determining the order status verification result based on the current order status; determining the order quantity verification result based on the delivery personnel order information; determining the location consistency verification result based on the delivery personnel real-time location information; and determining the data verification result based on the order status verification result, the order quantity verification result, and the location consistency verification result.

[0061] Specifically, by calling the getOrderInfo() interface of the Dify platform and passing in the order identifier, you can obtain the current order status and the shipping address information in the order. By passing in the delivery person identifier, you can obtain the delivery person's order information and the delivery person's real-time location information. The interface returns the current order status, delivery person's order information, and delivery person's real-time location information in the target test environment.

[0062] The system checks if the current order status matches the target status code. If yes, the order status verification is successful; otherwise, it fails. It determines the number of orders placed by the delivery personnel based on their order information. If the number is 0, it marks "no_order," indicating no orders, unavailable, and verification fails. If the number is greater than 1, it marks "multi_order," indicating multiple orders, which can cause AI context confusion, making it unavailable and verification fails. If the number is 1, the order quantity verification is successful. The system then determines the location distance between the delivery personnel and the sender based on the latitude and longitude of the delivery personnel's real-time information and the order's shipping address. If the location distance is greater than a preset distance threshold, the location consistency verification fails; if the location distance is not greater than the preset distance threshold, the location consistency verification passes.

[0063] The data verification result is determined based on the order status verification result, the order quantity verification result, and the location consistency verification result. Specifically, the data verification result is considered passed when all three verification results are valid. If any one of these verification results fails, the data verification result is considered failed.

[0064] If the data verification result is successful, test cases are generated based on the order identifier, delivery personnel identifier, and associated parameter information. If the data verification result is unsuccessful, an alarm is triggered, and testers intervene. The delivery personnel identifier (courier_id), order identifier (orderNumber), order latitude and longitude information from the associated parameter information, delivery personnel latitude and longitude information, and city identifier are used as `agent_inputs` to fill into the test case template, resulting in the test cases. During the actual testing phase, the test cases are imported into the AI ​​customer service evaluation platform running in the target testing environment. The AI ​​customer service evaluation platform initiates a session request, passing the `agent_inputs` from the test cases to the AI ​​customer service agent (Agent) of the generative AI application development platform, starting a simulated multi-round session. The AI ​​customer service evaluation platform can collect the evaluation reports output from each round to further determine the evaluation results of the instant delivery AI customer service.

[0065] For example, in a multi-turn AI customer service testing scenario, Dify simulates multi-turn dialogues as follows, taking a two-turn conversation as an example: Round 1: Receive user query "I want to change my mailing address", read the order identifier (DD_TEST_0088) in agent_inputs, call the internal getOrderInfo() interface, query the target test environment in real time for order DD_TEST_0088, and get the current order status as "44" (picked up); Agent performs business logic judgment, outputs customer service reply, and the server saves the session context.

[0066] Second round: Send in the user's second query "The new address is XX District XX Street", reuse the same session, continue reasoning, call the backend address change interface, and actually modify this order in the target test environment.

[0067] This embodiment's technical solution, upon determining the target state information as a target state code, determines the target state path based on a pre-constructed directed graph of order states. It then invokes the state progression tool for the target test environment, advancing the order state within the target test environment based on the target state path, order identifier, and delivery personnel identifier. This ensures that the order state within the target test environment corresponds to the order state code. Based on the order identifier, delivery personnel identifier, and associated parameter information, test cases are generated. This technical solution automatically generates the target state path and prepares real orders in specified business states in batches based on the target state code provided by the tester. This reduces manual memorization of state processes and step-by-step operation steps. State code planning and validity verification reduce data generation failures caused by illegal state transitions and ensure that order states conform to real business processes, thereby guaranteeing the accuracy of evaluations in multi-turn conversational AI customer service scenarios.

[0068] Example 3 Figure 3 This is a flowchart illustrating a method for generating evaluation test cases for AI-powered instant delivery customer service, as provided in Embodiment 3 of the present invention. This embodiment provides a preferred example based on the above embodiments.

[0069] like Figure 3 As shown, the method includes the following steps: S31. In response to the tester's request to generate test cases, parse the request to determine the target status information, test environment identifier, and associated parameter information.

[0070] S32. Determine the target test environment based on the test environment identifier, call the delivery personnel registration interface of the target test environment to construct the delivery personnel entity, and determine the delivery personnel identifier corresponding to the delivery personnel entity.

[0071] S33. The order placement test case of the call interface test platform is executed in the target test environment to obtain the test results, and the order identifier is determined according to the test results; if the target status information is the target status code, then execute S34; if the target status information is the initial status path, then execute S35.

[0072] S34. Based on the target status code and the pre-built directed graph of order status, determine the target status path.

[0073] S35. Validate the initial state path according to the directed graph of order status and determine whether the validation passes; if yes, proceed to S36; if no, proceed to S37-S39.

[0074] S36. Determine the initial state path as the target state path.

[0075] S37. Based on the directed graph of order status, the initial status path is corrected to obtain the corrected status path and error message.

[0076] S38. Feed back the corrected status path and error message to the testers so that they can confirm the error message and corrected status path and provide feedback on the confirmation result.

[0077] S39. If the confirmation result is a path confirmation result for the corrected state path, then the corrected state path shall be determined as the target state path.

[0078] S40. Invoke the status advancement tool of the target test environment. Based on the target status path, order identifier, and delivery personnel identifier, advance the order status within the target test environment to ensure that the order status within the target test environment is the order status corresponding to the target status code.

[0079] S41. Call the order information acquisition interface of the generative AI application development platform, and perform data verification in the target test environment based on the delivery personnel identifier and the order identifier to obtain the data verification result.

[0080] S42. Determine whether the data verification result is successful. If yes, execute S43A; otherwise, execute S43B.

[0081] S43A. Generate test cases based on order identifier, delivery personnel identifier, and associated parameter information.

[0082] S43B generates alarm notifications for testers.

[0083] S44. Import the test cases into the AI ​​customer service evaluation platform running in the target test environment. The AI ​​customer service evaluation platform initiates a session request, passes the agent_inputs in the test cases to the AI ​​customer service agent of the Dify platform, and starts a simulated multi-round session. The AI ​​customer service evaluation platform can collect the evaluation report output in each round to further determine the evaluation results of the instant delivery AI customer service.

[0084] In a specific implementation scenario, this method is illustrated using the preparation of evaluation data for an order in the "pickup" status as an example. The tester only provides the target status code target-code=44. First, the delivery personnel registration tool is called to register a test delivery personnel in the default status and retrieve the delivery personnel identifier courier_id; then, an order placement test case on the interface testing platform (e.g., the Open Platform V5 order placement test case running on the interface testing platform) is called, switching to the specified target testing environment, triggering asynchronous execution, and polling the report, extracting the real order identifier orderNumber from the returned response_body.

[0085] During target state path planning, a BFS shortest path search is performed on the directed graph of the order state, starting from the initial state of START (order placement). This yields an executable state code sequence [30, 40, 42, 44] leading to target 44 (if the target state code is 60, the planned state code sequence is [30, 40, 42, 44, 52, 60]). The order state is then progressively set using the state code sequence and the state progression tool. If any step fails, the order's progress is halted and the final state is recorded. For comparison, if the tester manually specifies the state code sequence chain-codes=[30, 52], each valid edge is checked from the initial state of START. If state 30 cannot directly transition to 52, an illegal step is identified and an alarm is triggered, resulting in rejection of execution or a prompt for correction.

[0086] During the precondition self-check phase, the `getOrderInfo` function of the Dify platform is called to verify the order status of the courier. Only orders with matching location coordinates and shipping addresses are considered usable evaluation data. Finally, the fields `courier_id`, `order_number`, latitude and longitude, and city identifier are populated back into the `conversation_turns.agent_inputs` field of the evaluation test case template, where `conversation_turns` represents the number of session rounds, allowing the evaluation test case to run directly. Order placement and status diagram planning are automatically linked within a single batch command, reducing manual switching between multiple tools and missed steps.

[0087] Example 4 Figure 4 This is a schematic diagram of the structure of an instant delivery AI customer service evaluation test case generation device provided in Embodiment 4 of the present invention. The instant delivery AI customer service evaluation test case generation device provided in this embodiment of the present invention is applicable to the automated generation of test cases for multi-turn dialogue AI customer service evaluation scenarios. This instant delivery AI customer service evaluation test case generation device can be implemented in hardware and / or software, such as... Figure 4 As shown, the device includes: a request response module 401, a personnel identification determination module 402, an order identification determination module 403, and an evaluation case generation module 404. Among them, The request response module 401 is used to respond to the tester's test case generation request, parse the test case generation request, and determine the target status information, test environment identifier, and associated parameter information; The personnel identification module 402 is used to determine the target test environment based on the test environment identifier, call the delivery personnel registration interface of the target test environment to construct a delivery personnel entity, and determine the delivery personnel identifier corresponding to the delivery personnel entity; The order identifier determination module 403 is used to call the order placement test cases of the interface test platform to execute in the target test environment, obtain the test results, and determine the order identifier based on the test results; The evaluation test case generation module 404 is used to generate evaluation test cases based on the target status information, the delivery personnel identifier, the order identifier, and the associated parameter information. The evaluation test cases are used to import into the AI ​​customer service evaluation platform running in the target test environment, and the AI ​​customer service intelligent agent of the generative AI application development platform is driven by the AI ​​customer service evaluation platform to perform real-time delivery AI customer service simulation evaluation.

[0088] The technical solution of this invention determines the target test environment based on the test environment identifier, constructs a delivery person entity by calling the delivery person registration interface of the target test environment, and determines the delivery person identifier corresponding to the delivery person entity; executes the order placement test case of the interface test platform in the target test environment to obtain the test result, and determines the order identifier based on the test result; and generates evaluation test cases based on the target status information, delivery person identifier, order identifier, and associated parameter information. The above technical solution realizes the automated generation of test cases in the instant delivery AI customer service evaluation scenario. By automatically generating data based on the status identifier given by the tester, it advances the order execution status, improves the efficiency of test case generation in the instant delivery AI customer service evaluation scenario, and improves the accuracy of test case generation in the instant delivery AI customer service evaluation scenario by eliminating the need for manual memorization of the chain.

[0089] Optionally, the test case generation module 404 includes: The first state path generation unit is used to determine the target state path based on the pre-constructed directed graph of order status, according to the target status code if the target state information is a target status code. The order status advancement unit is used to call the status advancement tool of the target test environment, and advance the order status within the target test environment based on the target status path, the order identifier, and the delivery personnel identifier, so as to ensure that the order status within the target test environment is the order status corresponding to the target status code; The first test case generation unit is used to generate test cases based on the order identifier, the delivery personnel identifier, and the associated parameter information.

[0090] Optionally, the test case generation module 404 also includes: The data verification unit is used to call the order information acquisition interface of the generative AI application development platform before generating test cases based on the order identifier, the delivery personnel identifier and the associated parameter information, and to perform data verification in the target test environment based on the delivery personnel identifier and the order identifier to obtain the data verification result; The second test case generation unit is used to generate test cases based on the order identifier, the delivery personnel identifier, and the associated parameter information if the data verification result is successful.

[0091] Optionally, the test case generation module 404 also includes: A path verification unit is used to verify the initial state path according to the directed graph of the order state if the target state information is an initial state path. The second state path determination unit is used to determine the initial state path as the target state path if the verification passes.

[0092] Optionally, the test case generation module 404 also includes: The path correction unit is used to correct the initial state path based on the directed graph of the order status if the verification fails, so as to obtain the corrected state path and error message. The prompt information feedback unit is used to feed back the corrected status path and the error prompt information to the tester, so that the tester can confirm the error prompt information and the corrected status path and provide feedback on the confirmation result; The third state path determination unit is used to determine the corrected state path as the target state path if the confirmation result is a path confirmation result for the corrected state path.

[0093] Optionally, the directed graph of order status can be constructed as follows: Determine several order execution states in the context of instant order delivery, and construct status codes corresponding to each of the order execution states; The status code corresponding to the execution status of each order is determined as a status node; Based on the association between the execution states of each order, construct state node edges between the state nodes; Generate metadata for the edges of the state nodes; the metadata includes the starting state code, the target state code, the preconditions for the transition, the transition dependency parameter field, the idempotent key, and the information on the compensation operation for the transition failure; Construct a directed graph of order states that includes each of the aforementioned state nodes and their corresponding state node edges.

[0094] Optional, a data verification unit, specifically used for: The order information acquisition interface of the generative AI application development platform is called, and the order identifier and the delivery personnel identifier are used as interface parameters to obtain the current order status, delivery personnel order information and delivery personnel real-time location information in the target test environment; Determine the order status verification result based on the current order status; and, The order quantity verification result is determined based on the order information of the delivery personnel. The location consistency verification result is determined based on the real-time location information of the delivery personnel. The data verification result is determined based on the order status verification result, the pending order quantity verification result, and the location consistency verification result.

[0095] The instant delivery AI customer service evaluation test case generation device provided in this embodiment of the invention can execute the instant delivery AI customer service evaluation test case generation method provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0096] Example 5 Figure 5 A schematic diagram of an electronic device 50 that can be used to implement embodiments of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices (e.g., helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.

[0097] like Figure 5 As shown, the electronic device 50 includes at least one processor 51 and a memory, such as a read-only memory (ROM) 52 and a random access memory (RAM) 53, communicatively connected to the at least one processor 51. The memory stores computer programs executable by the at least one processor. The processor 51 can perform various appropriate actions and processes based on the computer program stored in the ROM 52 or loaded from storage unit 58 into the RAM 53. The RAM 53 can also store various programs and data required for the operation of the electronic device 50. The processor 51, ROM 52, and RAM 53 are interconnected via a bus 54. An input / output (I / O) interface 55 is also connected to the bus 54.

[0098] Multiple components in electronic device 50 are connected to I / O interface 55, including: input unit 56, such as keyboard, mouse, etc.; output unit 57, such as various types of monitors, speakers, etc.; storage unit 58, such as disk, optical disk, etc.; and communication unit 59, such as network card, modem, wireless transceiver, etc. Communication unit 59 allows electronic device 50 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0099] Processor 51 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 51 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, digital signal processors (DSPs), and any suitable processor, controller, microcontroller, etc. Processor 51 performs the various methods and processes described above, such as the instant delivery AI customer service evaluation use case generation method.

[0100] In some embodiments, the instant delivery AI customer service evaluation use case generation method can be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 58. In some embodiments, part or all of the computer program can be loaded and / or installed on electronic device 50 via ROM 52 and / or communication unit 59. When the computer program is loaded into RAM 53 and executed by processor 51, one or more steps of the instant delivery AI customer service evaluation use case generation method described above can be performed. Alternatively, in other embodiments, processor 51 can be configured to execute the instant delivery AI customer service evaluation use case generation method by any other suitable means (e.g., by means of firmware).

[0101] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), payload-programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0102] Computer programs used to implement the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the processor, the computer programs cause the functions / operations specified in the flowcharts and / or block diagrams to be performed. The computer programs may be executed entirely on a machine, partially on a machine, or as a standalone software package, partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0103] In the context of this invention, a computer-readable storage medium can be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, apparatus, or device. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination thereof. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0104] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the electronic device. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0105] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or middleware components (e.g., application servers), or frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include local area networks (LANs), wide area networks (WANs), blockchain networks, and the Internet.

[0106] A computing system can include clients and servers. Clients and servers are generally located far apart and typically interact through communication networks. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a hosting product within the cloud computing service system to address the shortcomings of traditional physical hosts and VPS services, such as high management difficulty and weak business scalability.

[0107] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this invention can be achieved, and this is not limited herein.

[0108] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A method for generating evaluation test cases for AI customer service in instant delivery, characterized in that, include: In response to the tester's request to generate test cases, the request is parsed to determine the target status information, test environment identifier, and associated parameter information; The target test environment is determined based on the test environment identifier. The delivery personnel registration interface of the target test environment is called to construct the delivery personnel entity and the delivery personnel identifier corresponding to the delivery personnel entity is determined. The order placement test cases of the API testing platform are executed in the target testing environment to obtain test results, and the order identifier is determined based on the test results. Based on the target status information, the delivery personnel identifier, the order identifier, and the associated parameter information, evaluation cases are generated. The evaluation cases are imported into the AI ​​customer service evaluation platform running in the target test environment, and the AI ​​customer service agent of the generative AI application development platform is driven by the AI ​​customer service evaluation platform to perform real-time delivery AI customer service simulation evaluation.

2. The method according to claim 1, characterized in that, The step of generating evaluation test cases based on the target status information, the delivery personnel identifier, the order identifier, and the associated parameter information includes: If the target status information is a target status code, then the target status path is determined based on the target status code and the pre-constructed directed graph of order status. The state advancement tool of the target test environment is invoked to advance the order state within the target test environment based on the target state path, the order identifier, and the delivery personnel identifier, so as to ensure that the order state within the target test environment is the order state corresponding to the target state code; Evaluation cases are generated based on the order identifier, the delivery personnel identifier, and the associated parameter information.

3. The method according to claim 2, characterized in that, Before generating test cases based on the order identifier, the delivery person identifier, and the associated parameter information, the method further includes: The order information acquisition interface of the generative AI application development platform is called to perform data verification in the target test environment based on the delivery personnel identifier and the order identifier, and the data verification result is obtained. If the data verification result is successful, then the step of generating evaluation cases based on the order identifier, the delivery personnel identifier, and the associated parameter information is executed.

4. The method according to claim 2, characterized in that, The method further includes: If the target state information is an initial state path, then the initial state path is verified according to the directed graph of the order state. If the verification passes, the initial state path will be determined as the target state path.

5. The method according to claim 4, characterized in that, The method further includes: If the verification fails, the initial state path is corrected based on the directed graph of the order status to obtain the corrected state path and error message. The corrected status path and the error message are fed back to the tester so that the tester can confirm the error message and the corrected status path and provide feedback on the confirmation result. If the confirmation result is a path confirmation result for the corrected state path, then the corrected state path is determined as the target state path.

6. The method according to claim 2, characterized in that, The directed graph of order status is constructed as follows: Determine several order execution states in the context of instant order delivery, and construct status codes corresponding to each of the order execution states; The status code corresponding to the execution status of each order is determined as a status node; Based on the association between the execution states of each order, construct state node edges between the state nodes; Generate metadata for the edges of the state nodes; the metadata includes the starting state code, the target state code, the preconditions for the transition, the transition dependency parameter field, the idempotent key, and the information on the compensation operation for the transition failure; Construct a directed graph of order states that includes each of the aforementioned state nodes and their corresponding state node edges.

7. The method according to claim 3, characterized in that, The step of calling the order information acquisition interface of the generative AI application development platform, performing data verification within the target test environment based on the delivery personnel identifier and the order identifier, and obtaining the data verification result includes: The order information acquisition interface of the generative AI application development platform is called, and the order identifier and the delivery personnel identifier are used as interface parameters to obtain the current order status, delivery personnel order information and delivery personnel real-time location information in the target test environment; Determine the order status verification result based on the current order status; and, The order quantity verification result is determined based on the order information of the delivery personnel. The location consistency verification result is determined based on the real-time location information of the delivery personnel. The data verification result is determined based on the order status verification result, the pending order quantity verification result, and the location consistency verification result.

8. A device for generating evaluation test cases for AI-powered instant delivery customer service, characterized in that, include: The request and response module is used to respond to the testers' request to generate test cases, parse the request to determine the target status information, test environment identifier and associated parameter information; The personnel identification module is used to determine the target test environment based on the test environment identifier, call the delivery personnel registration interface of the target test environment to construct a delivery personnel entity, and determine the delivery personnel identifier corresponding to the delivery personnel entity; The order identifier determination module is used to call the order placement test cases of the interface test platform to execute in the target test environment, obtain the test results, and determine the order identifier based on the test results; The evaluation test case generation module is used to generate evaluation test cases based on the target status information, the delivery personnel identifier, the order identifier, and the associated parameter information. The evaluation test cases are imported into the AI ​​customer service evaluation platform running in the target test environment, and the AI ​​customer service intelligent agent of the generative AI application development platform is driven by the AI ​​customer service evaluation platform to perform real-time delivery AI customer service simulation evaluation.

9. An electronic device, characterized in that, The electronic device includes: At least one processor; and A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the instant delivery AI customer service evaluation use case generation method according to any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions that cause a processor to execute the instant delivery AI customer service evaluation test case generation method according to any one of claims 1-7.

Images