Dialogue type catering coupon delivery system based on multiple agents

The conversational restaurant coupon distribution system with a multi-agent architecture solves the problems of passive interaction, isolated decision-making, and fragmented execution in traditional systems, achieving efficient coupon generation and automated execution, and improving the ease of operation and strategy quality for merchants.

CN120912263APending Publication Date: 2025-11-07ACEWILL INFORMATION TECH BEIJING CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202511013244.8
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2025-07-23
Publication Date
2025-11-07

AI Technical Summary

Technical Problem

Traditional restaurant voucher distribution systems suffer from passive interaction, isolated decision-making, and fragmented execution, resulting in low efficiency in demand confirmation, high risk of strategy deviation, and low execution efficiency.

Method used

Employing a multi-agent architecture, the system achieves conversational interaction, dynamic data capture, and feedback guidance through task decomposition and collaborative reasoning. Combined with automated execution of multi-source data, it constructs an end-to-end closed-loop system, including modules for task determination, system planning, and task execution.

Benefits of technology

It enhances the depth of human-machine collaboration, saves time in the coupon issuance process, improves coupon quality and generation efficiency, and allows merchants to query database information and quickly analyze data at any time.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN120912263A_ABST
    Figure CN120912263A_ABST
Patent Text Reader

Abstract

The invention discloses a dialogue type catering coupon delivery system based on multiple agents, and relates to the technical field of data analysis, and specifically, the system comprises a task determination module which is used for making a dialogue with a merchant, guiding the merchant to perfect coupon delivery information and confirming task information; the system planning module is used for counting the number of ticket issuing target groups meeting the task information and membership card numbers of the ticket issuing target groups, and automatically executing ticket issuing planning for commercial tenants to refer to; and the task execution module is used for confirming the coupon issuing information with the merchant, and calling the coupon issuing interface to execute the coupon issuing operation if the merchant confirms the coupon issuing information. According to the invention, through task decomposition and collaborative reasoning among intelligent agents, dynamic capture and feedback guidance of the system on user intentions and automatic execution arrangement of multi-source data, an end-to-end closed loop from demand conversation to strategy generation and from data linkage to execution triggering is realized, and the man-machine collaborative depth and strategy generation efficiency are effectively improved.
Need to check novelty before this filing date? Find Prior Art

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data analysis, in particular to a multi-agent based conversational dining coupon issuing system. BACKGROUND

[0002] With the acceleration of the digitalization process of the catering industry, online marketing has become the core means for merchants to obtain customer flow. However, the traditional coupon issuing mode generally has problems such as passive interaction, isolated decision-making, and fragmented execution. On the one hand, merchants need to repeatedly confirm the coupon issuing demand through complex forms or manual communication, resulting in a semantic gap between demand expression and actual operation goals. On the other hand, the design of the discount strategy is limited to a static rule engine, making it difficult to dynamically adjust in combination with multi-dimensional indicators such as redemption rate, average order value, and time period customer flow.

[0003] The current dining coupon issuing system faces three technical barriers: First, the interaction layer lacks a guided dialogue mechanism, and merchants need to repeatedly switch between multiple interfaces and are difficult to obtain strategy suggestions, resulting in low efficiency of demand confirmation.

[0004] Existing dining coupon issuing systems generally use a form filling or fixed dialogue question and answer interaction mode. Merchants need to manually switch between multiple interfaces to complete the coupon issuing information, and the system lacks active guidance ability. This static interaction method cannot dynamically capture the real business demands of merchants (such as increasing average order value, attracting idle time customer flow, etc.) through dialogue, resulting in a lengthy demand collection process and easy omission of key parameters. Ultimately, the discount strategy generated often deviates from the actual operation goals.

[0005] Second, the decision-making layer is limited by single-point data analysis capability and cannot dynamically adapt the discount plan to the business goals through multi-agent collaboration, which easily leads to strategy deviation risk.

[0006] Traditional systems rely on a single rule engine or simple machine learning model to generate discount plans, which can only make limited adjustments based on pre-set templates and cannot combine multi-source data such as redemption rate, average order value, and time period customer flow for multi-objective optimization. At the same time, existing solutions lack task decomposition and collaborative reasoning mechanisms among multiple agents, making it difficult to simulate the decision-making logic of professional marketing consultants (such as balancing discount intensity and gross margin space), resulting in discount strategies easily falling into local optimal solutions.

[0007] Third, the execution layer involves the integration of heterogeneous data sources such as POS systems, delivery platform APIs, and CRM databases, and traditional monolithic architecture has significant shortcomings in parameter configuration accuracy and cross-platform synchronization timeliness.

[0008] The coupon issuing needs to be connected with POS systems, take-out platform APIs, CRM databases and other heterogeneous data sources. However, the existing systems mostly adopt monolithic architecture, and cannot realize automatic configuration and real-time synchronization of cross-platform parameters. Manual intervention in the execution link is easy to cause configuration errors (such as time interval dislocation and use threshold conflict), and lacks dynamic response capability to abnormal states (such as sudden high concurrency redemption), resulting in low execution efficiency and poor risk controllability.

[0009] In this context, the new generation of intelligent systems that integrate multi-agent collaboration and conversational interaction have become the technical direction to break through the industry bottleneck. SUMMARY

[0010] The existing technology has problems such as passive interaction, isolated decision-making and fragmented execution in the traditional coupon distribution mode. In order to solve the above technical problems, the technical scheme of the present application is as follows: A dialog-based restaurant coupon distribution system based on multi-agent, comprising A task determination module, which is used for dialog with a merchant, guiding the merchant to perfect coupon distribution information, and confirming task information; A system planning module, which is used for counting the number of coupon distribution target groups and member card numbers of the coupon distribution target groups that meet the task information, and automatically executing coupon distribution planning for the merchant to refer to; An execution task module, which is used for confirming coupon distribution information with the merchant, and if the merchant finds that the coupon distribution information needs to be adjusted, the task determination module is jumped to to reconfirm the task information after the merchant provides feedback, and if the merchant confirms the coupon distribution information, a coupon distribution interface is called to execute the coupon distribution operation.

[0011] Preferably, the coupon distribution information includes the validity period of the coupon, the starting date of the coupon and the coupon discount rules.

[0012] Preferably, the task determination module includes a dialog module, an information library and a coupon distribution suggestion module, the dialog module searches the information library according to the requirements of the merchant, and dialogues with the merchant while calling the coupon distribution suggestion module to give the merchant coupon distribution suggestions.

[0013] Preferably, the information library includes a coupon information library and a merchant member information library.

[0014] Preferably, the task determination module further includes an agent layer, the agent layer is connected with the coupon distribution suggestion module, the agent layer includes an information analysis agent and a networked information search agent, the information analysis agent is used for analyzing historical coupon distribution information and member information for the merchant to refer to, and the networked search agent is used for searching the whole network historical coupon distribution experience for the merchant to refer to.

[0015] Preferably, the information missing discrimination agent is further included for distinguishing whether the coupon information is missing and reminding the merchant to improve.

[0016] Preferably, the system planning module includes a data search module and a data analysis module, the data search module searches and gives a plan according to the data range given by the merchant, and the data analysis module performs refined analysis according to the dialogue information with the merchant and gives data display.

[0017] Preferably, the execution task module includes an automatic coupon issuing module and an information confirmation module, the information confirmation module is used to confirm the coupon information with the merchant, if the merchant confirms the coupon information, the automatic coupon issuing module is called to execute the coupon issuing operation, if the merchant finds that the step has an error and needs to adjust the information, it is fed back to the task confirmation module.

[0018] Compared with the prior art, the present application has the following beneficial effects: (1) The dialogue type multi-agent architecture of the dialogue type catering coupon issuing system based on multi-agent proposed by the present application realizes end-to-end closed loop from demand dialogue to strategy generation, data linkage to execution triggering through task decomposition and collaborative reasoning between agents, dynamic capture and feedback guidance of system to user intention, and automatic execution arrangement of multi-source data, effectively improving the depth of human-computer collaboration and the efficiency of strategy generation.

[0019] (2) The dialogue type catering coupon issuing system based on multi-agent guides the merchant to formulate reasonable coupons and executes automatic coupon issuing, which effectively saves the time consumption of the coupon issuing process and effectively improves the quality of the coupons. Moreover, the merchant can query the information of the self database at any time, can quickly search and analyze the data of the self database, and is convenient and fast to use. BRIEF DESCRIPTION OF DRAWINGS

[0020] Figure 1 The system block diagram of the dialogue type catering coupon issuing system based on multi-agent of the present application; Figure 2 The functional flow chart of the dialogue type catering coupon issuing system based on multi-agent of the present application; Figure 3 The information technology flow chart for determining task details in the dialogue type catering coupon issuing system based on multi-agent of the present application; Figure 4 The information technology flow chart for planning task steps in the dialogue type catering coupon issuing system based on multi-agent of the present application; Figure 5 The information technology flow chart for executing task steps in the dialogue type catering coupon issuing system based on multi-agent of the present application; Figure 6An interface diagram of a multi-agent-based conversational meal coupon issuing system according to the present application. DETAILED DESCRIPTION

[0021] The specific embodiments of the present application will be further described below with reference to the drawings. It should be noted that the description of these embodiments is used to help understand the present application, but does not constitute a limitation on the present application. In addition, the technical features involved in the various embodiments of the present application described below can be combined with each other as long as they do not conflict with each other.

[0022] As Figures 1 to 6 shown, the present application discloses a multi-agent-based conversational meal coupon issuing system. The system uses the langgraph framework to build a graph structure and realizes the function flow through node jumping. The task is divided into three stages: task detail confirmation, task plan formulation, and task plan execution. Complex tasks must go through task detail confirmation and task plan formulation. In the task detail confirmation stage, if it is a coupon issuing task, the system will dialogue with the merchant to guide the merchant to perfect the coupon issuing detail information. In addition, a coupon issuing suggestion node is added for the coupon issuing task. The node is mounted with two agents, including an information analysis agent that analyzes historical coupon issuing information and member information, and a network search agent that searches for the relevant coupon issuing experience of listed companies. The node will cooperate with the task information confirmation node to guide the merchant to formulate a reasonable and effective coupon through continuous dialogue with the merchant. In the task plan formulation stage, the planning node will split the task into steps, and each step will correspond to the use of a certain agent node. In the task execution stage, the plan execution control node will read the plan and control the sequential execution of the plan. Different tasks will be completed sequentially by combining multiple different agent nodes.

[0023] Specifically, a multi-agent-based conversational meal coupon issuing system includes the following modules, A task determination module, as Figure 3 shown, the task determination module is used to dialogue with the merchant to guide the merchant to perfect the coupon issuing information and confirm the task information. The specific coupon issuing information includes the coupon validity period, the coupon starting date, the coupon discount rules, etc. The task determination module includes a dialogue module, an information library, and a coupon issuing suggestion module. The dialogue module searches the information library according to the requirements of the merchant and dialogues with the merchant, while the coupon issuing suggestion module is called to give the merchant coupon issuing suggestions. The information library includes a coupon information library and a merchant member information library, as well as other information knowledge bases required for the task, such as coupon issuing tasks, data analysis tasks, network search tasks, etc.

[0024] As Figure 3As shown, preferably, the task determination module also includes an intelligent agent layer, which is connected to the coupon issuance suggestion module. The intelligent agent layer includes an information analysis intelligent agent and a network information search intelligent agent. The information analysis intelligent agent analyzes historical coupon issuance information and member information for merchants' reference, while the network search intelligent agent searches the entire network for historical coupon issuance experience for merchants' reference. During the dialogue between the merchant and the system, the system combines network search data and past coupon issuance experience to provide merchants with relevant coupon issuance suggestions. For example, if a merchant wants to issue coupons for the Dragon Boat Festival, the system will search the network for the Dragon Boat Festival marketing activities of listed catering companies and the merchant's historical Dragon Boat Festival coupon issuance. Combining these two pieces of information, the system will provide feedback to the merchant, offering design suggestions for Dragon Boat Festival coupons, thereby guiding the merchant to optimize the details of setting coupons until the task information is determined.

[0025] like Figure 3 As shown, in a further preferred embodiment, the task determination module also includes an information missing detection intelligent agent, which is used to distinguish whether the coupon issuance information is missing and remind the merchant to complete it, so as to facilitate the confirmation of task information.

[0026] System planning module, such as Figure 4 As shown, the system planning module is used to count the number of people in the target group for coupon issuance and their membership card numbers, and automatically executes the coupon issuance plan for merchants' reference. The system planning module includes a data search module and a data analysis module. The data search module searches based on the data range provided by the merchant and provides a plan. The data analysis module performs detailed analysis based on the dialogue information with the merchant and provides data visualization. Regarding the data search module, for example, if a merchant needs to find the number of customers who have made purchases at a certain store in the past three months, first, the merchant submits a request to the system. Then, the data search module filters the data according to the request. Once the system confirms that the information required to execute the task is complete, it plans the query operation and finally executes the query. Regarding the data analysis module, for example, if a merchant needs to analyze the consumer group of a certain store of a certain merchant in the past three months, first, the merchant communicates with the system to confirm whether the analysis needs to be further refined, such as adding specific analysis group scopes. After the merchant confirms the relevant analysis objectives, the planning phase begins. The data analysis module plans the analysis process as follows: Step 1, count the number of consumers in the past 3 months; Step 2, perform group statistics based on the existing group profile information; Step 3, draw the statistical results chart; Step 4, summarize the statistical results for customer reference.

[0027] Execution task module, such as Figure 5As shown, the task execution module is used to confirm the coupon issuing information with the merchant. If the merchant finds that the coupon issuing information needs to be adjusted, the merchant provides feedback, and then jumps to the task determination module to reconfirm the task information. If the merchant confirms the coupon issuing information, the coupon issuing interface is called to execute the coupon issuing operation. Specifically, the task execution module includes an automatic coupon issuing module and an information confirmation module. The information confirmation module is used to confirm the coupon issuing information with the merchant. If the merchant confirms the coupon issuing information, the automatic coupon issuing module is called to execute the coupon issuing operation. If the merchant finds that an error occurs in the step and needs to adjust the information, the feedback is provided to the task confirmation module.

[0028] The embodiments of the present application are described in detail above with reference to the drawings, but the present application is not limited to the described embodiments. For those skilled in the art, various changes, modifications, replacements and variations of the embodiments can be made without departing from the principles and spirits of the present application, and still fall within the protection scope of the present application.

Claims

1. A multi-agent based conversational voucher distribution system, characterized in that: The system comprises a task determination module, a system planning module, and an execution task module. The task determination module is used to dialogue with the merchant, guide the merchant to perfect the coupon issuing information, and confirm the task information. The system planning module is used to count the number of the coupon issuing target groups and the member card numbers of the coupon issuing target groups that meet the task information, and automatically execute the coupon issuing plan for the merchant to refer to.

2. The multi-agent based conversational voucher distribution system according to claim 1, wherein: The execution task module is used to confirm the coupon issuing information with the merchant.

3. The multi-agent based conversational voucher distribution system according to claim 1, wherein: If the merchant finds that the coupon issuing information needs to be adjusted, the system jumps to the task determination module to reconfirm the task information after the merchant gives feedback.

4. The multi-agent based conversational voucher distribution system according to claim 3, wherein: If the merchant confirms the coupon issuing information, the system calls the coupon issuing interface to execute the coupon issuing operation.

5. The multi-agent based conversational voucher distribution system according to claim 3, wherein: The coupon issuing information comprises the coupon validity period, the coupon starting date, and the coupon preferential rules.

6. The multi-agent based conversational voucher dispensing system according to claim 5, wherein: The task determination module comprises a dialogue module, an information database, and a coupon issuing suggestion module.

7. The multi-agent based conversational voucher delivery system according to claim 1, wherein: The dialogue module searches the information database according to the requirements of the merchant, dialogues with the merchant, and calls the coupon issuing suggestion module to give the coupon issuing suggestions to the merchant. 8.The multi-agent based conversational voucher delivery system of claim 1, wherein: The information database comprises a coupon information database and a merchant member information database. The task determination module further comprises an agent layer connected with the coupon issuing suggestion module. The agent layer comprises an information analysis agent and a networked information search agent. The information analysis agent is used to analyze the historical coupon issuing information and the member information for the merchant to refer to. The networked information search agent is used to search the historical coupon issuing experience of the whole network for the merchant to refer to. The system further comprises an information missing discrimination agent. The information missing discrimination agent is used to distinguish whether the coupon issuing information is missing, and remind the merchant to perfect the information. The system planning module comprises a data search module and a data analysis module. The data search module searches according to the data range given by the merchant and gives the plan. The data analysis module performs detailed analysis according to the dialogue information with the merchant and gives the data display. The execution task module comprises an automatic coupon issuing module and an information confirmation module. The information confirmation module is used to confirm the coupon issuing information with the merchant. If the merchant confirms the coupon issuing information, the system calls the automatic coupon issuing module to execute the coupon issuing operation. If the merchant finds that the steps are wrong and the information needs to be adjusted, the system feeds back to the task confirmation module.