A decision engine-based loan review and dispatch method, server and storage medium

Abstract

The application provides a loan review dispatching method based on a decision engine, a server and a storage medium. The loan review dispatching method based on the decision engine comprises the following steps: in response to obtaining loan review information, extracting customer basic information and order information in the loan review information; calling the decision engine for strategy matching according to the customer basic information and the order information, so as to correspondingly obtain a customer risk classification code and an order risk classification code; calling the decision engine for type matching according to the customer risk classification code and the order risk classification code, so as to obtain a loan review type corresponding to the loan review information; and calling the decision engine for dispatching matching according to the loan review type, so as to determine a loan review officer corresponding to the loan review information. The loan review dispatching method based on the decision engine, the server and the storage medium provided by the application can perform differentiated loan review, improve the efficiency of loan review, reduce the cost of loan review, and provide better services for customers.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of financial credit audit, in particular to a loan audit assignment method based on a decision engine, a server and a storage medium. BACKGROUND

[0002] In the process of customer loan qualification audit in the field of financial credit, the basic situation of the customer, the credit investigation situation, the asset situation, the engineering contract, the purpose of the proposed equipment, the financial plan, the guarantee situation and all other fields or part of the fields are usually investigated and understood, and the risk situation of the applicant is comprehensively evaluated by accessing the third-party credit investigation data source to determine whether to provide credit loans. The traditional financing leasing company investigates the credit qualification of the customer, which is based on the company's own scale, product characteristics and other factors to take one of the credit audit methods such as loan audit by loan audit officers, customer online self-loan audit and remote video loan audit.

[0003] In the process of conceiving and implementing the present application, the inventors found that at least the following problems exist: the customer self-loan audit method only makes judgments based on the loan audit information submitted by the customer and the internal and external credit investigation situation, and the risk control of the customer's qualification is limited, which can only meet the small orders below the medium scale of the loan amount. The remote video loan audit method uses a video call method to dialogue with the customer, which improves the control of the customer's qualification and risk compared to the customer self-loan audit, but the authenticity of the customer's family situation, asset situation and purpose of the proposed equipment is still insufficient, which can only meet the small and medium orders below the medium scale of the loan amount. The loan audit officer on-site loan audit method uses the method of loan audit officers going to the customer site and the engineering site to understand the actual situation of the customer and control the risk, but when it comes to small orders and remote areas where the customer is located, the loan audit cost is high and the cost-effectiveness is low. In addition, due to some special control factors and other special factors, it is impossible to go to the customer site for on-site loan audit, and part of the business is lost.

[0004] The foregoing discussion is presented solely to provide general background information and does not necessarily constitute prior art. SUMMARY

[0005] In order to alleviate the above problems, the present application provides a loan audit assignment method based on a decision engine, a server and a storage medium.

[0006] In one aspect, the present application provides a loan audit assignment method based on a decision engine, specifically comprising:

[0007] In response to obtaining loan audit information, extracting customer basic information and order information in the loan audit information;

[0008] According to the customer basic information and the order information, calling a decision engine for strategy matching to correspondingly obtain customer risk classification codes and order risk classification codes;

[0009] According to the customer risk classification code and the order risk classification code, a decision engine is called for type matching to obtain a loan review type corresponding to the loan review information;

[0010] According to the loan review type, a decision engine is called for order matching to determine a loan review officer corresponding to the loan review information.

[0011] Optionally, the loan review order dispatching method based on the decision engine includes the following steps in the step of calling the decision engine for strategy matching according to the customer basic information and the order information to respectively obtain the customer risk classification code and the order risk classification code:

[0012] According to the customer ID number of the customer basic information, a customer credit record is queried;

[0013] A customer risk classification strategy set in the decision engine is called, and rule matching of the risk strategy set is performed according to the customer credit record to obtain a matched customer risk classification code;

[0014] The customer credit record includes at least one of customer type, region, external credit, asset condition, internal credit of the enterprise, historical purchase record and historical repayment information of the enterprise.

[0015] Optionally, the loan review order dispatching method based on the decision engine includes the following steps in the step of calling the decision engine for strategy matching according to the customer basic information and the order information to respectively obtain the customer risk classification code and the order risk classification code:

[0016] The credit records of historical customers are read, and first weight values are respectively assigned according to first weight dimensions of the historical customers, the first weight dimensions including customer type, region, external credit, asset condition, internal credit of the enterprise, historical purchase record and historical repayment information of the enterprise;

[0017] According to the first weight values, different customers are classified into a plurality of risk classifications, and customer risk classification codes are assigned to the plurality of risk classifications;

[0018] According to the first weight dimensions of the customer credit record of the loan review information, a risk classification code corresponding to the loan review information is determined.

[0019] Optionally, the loan review order dispatching method based on the decision engine includes the following steps in the step of calling the decision engine for strategy matching according to the customer basic information and the order information to respectively obtain the customer risk classification code and the order risk classification code:

[0020] According to the procurement financial scheme of the order information, the order information is subjected to rule matching of a credit policy strategy set in the decision engine to obtain a matched order risk classification code.

[0021] The procurement financial scheme includes at least one of a procurement equipment type, an order amount, a down payment ratio, a financial scheme type, and a guarantee condition.

[0022] Optionally, the loan review and order dispatching method based on the decision engine includes the following steps before the step of performing the rule matching of the credit policy strategy set in the decision engine according to the procurement financial scheme of the order information to obtain a matched order risk classification code.

[0023] The historical customer's procurement financial scheme is read, and a second weight dimension including a procurement equipment type, an order amount, a down payment ratio, a financial scheme type, and a guarantee condition is subjected to second weight assignment respectively.

[0024] According to the second weight assignment, different orders are classified into a plurality of risk classifications, and the plurality of risk classifications are subjected to order risk classification coding.

[0025] According to the second weight dimension of the procurement financial scheme of the order information, an order risk classification code corresponding to the order information is determined.

[0026] Optionally, the loan review and order dispatching method based on the decision engine includes the following steps before the step of performing the type matching by calling the decision engine according to the customer risk classification code and the order risk classification code to obtain a loan review type corresponding to the loan review information.

[0027] The scoring card model is called, and the customer risk classification code and the order risk classification code are inputted to make the scoring card model output a score for the loan review information.

[0028] The loan review information and the score are read, and rule matching is performed using a loan review type strategy set in the decision engine to determine a loan review type corresponding to the loan review information.

[0029] Optionally, the loan review type in the loan review and order dispatching method based on the decision engine includes at least one of a customer self-determined loan review, a remote video loan review, a loan review officer on-site loan review, and a double on-site loan review.

[0030] Optionally, the loan review and order dispatching method based on the decision engine includes the following steps before the step of performing the dispatching matching by calling the decision engine according to the loan review type to determine a loan review officer corresponding to the loan review information.

[0031] The decision tree model is called, the loan review information is input, and a credit manager qualification requirement corresponding to the loan review information is acquired;

[0032] The current workload of the credit manager meeting the credit manager qualification requirement is called, and the credit manager with the least current workload is determined as the loan review officer corresponding to the loan review information.

[0033] In another aspect, the application also provides a server, specifically, the server comprises a processor and a storage medium connected to each other, wherein:

[0034] The storage medium is used to store a computer program;

[0035] The processor is used to read the computer program and run, so that the server implements the loan review assignment method based on the decision engine as described above.

[0036] In another aspect, the application also provides a storage medium, specifically, the storage medium stores a computer program, and the computer program is executed by a processor to implement the steps of the loan review assignment method based on the decision engine as described above.

[0037] As described above, the loan review assignment method based on the decision engine, the server and the storage medium provided by the application automatically match the most suitable loan review type and the most optimal loan review officer responsible for decision making through multiple calls to the decision engine according to different dimensional attributes such as customers and orders, perform differentiated loan review, improve the efficiency of loan review, reduce loan review costs, and provide better services for customers. BRIEF DESCRIPTION OF DRAWINGS

[0038] The accompanying drawings incorporated into the specification and forming a part of the specification, show embodiments consistent with the application, and together with the specification, serve to explain the principles of the application. In order to more clearly illustrate the technical solutions of the embodiments of the application, the drawings needed in the embodiment description will be briefly introduced as follows. Obviously, for those of ordinary skill in the art, other drawings can also be obtained without creative labor based on these drawings.

[0039] Figure 1 Flow of the loan review assignment method based on the decision engine of an embodiment of the application Figure One .

[0040] Figure 2 Flow of the loan review assignment method based on the decision engine of an embodiment of the application Figure Two .

[0041] The objectives, features and advantages of the present application will be further illustrated in conjunction with the embodiments, with reference to the accompanying drawings. The above-mentioned drawings have shown the specific embodiments of the present application, which will be described in more detail hereinafter. These drawings and the written description are not intended to restrict the scope of the present application in any way, but to illustrate the concept of the present application for the technical personnel in the field by referring to the specific embodiments. DETAILED DESCRIPTION

[0042] The exemplary embodiments will be described in detail herein with reference to the accompanying drawings. In the following description, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments are not meant to represent all embodiments consistent with the present application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of the present application as detailed in the appended claims.

[0043] It should be noted that, in this document, the terms "comprises", "comprising", or any other variation thereof are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but can also include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element preceded by "comprises a" does not, without more constraints, foreclose the existence of additional identical elements in the process, method, article, or apparatus that comprises the element. Also, the same element denoted by the same reference numerals in different embodiments can have the same meaning or different meanings, which should be determined in light of the explanation of the element in the specific embodiment or further in light of the context in the specific embodiment.

[0044] It should be understood that the specific embodiments described herein are merely exemplary of the present application and are not intended to limit the present application in any way.

[0045] First Embodiment

[0046] In one aspect, the present application provides a decision engine based loan review assignment method, Figure 1 The flow of the decision engine based loan review assignment method of an embodiment of the present application Figure One , Figure 2 The flow of the decision engine based loan review assignment method of an embodiment of the present application Figure Two .

[0047] Please refer to Figure 1 and Figure 2 In one embodiment, the decision engine based loan review assignment method comprises:

[0048] S10: In response to obtaining the loan review information, extracting the customer basic information and order information in the loan review information.

[0049] When the customer needs a loan, the loan review information including the customer basic information and order information will be submitted.

[0050] S20: According to the customer basic information and order information, calling the decision engine for strategy matching to correspond to obtain the customer risk classification code and order risk classification code respectively.

[0051] Exemplarily, the decision engine includes various strategies required for intelligent order distribution, such as customer risk classification strategy, credit policy strategy, loan review type strategy, order distribution strategy, etc. The decision engine can quantitatively classify the customer risk and order risk through corresponding matching of the customer risk classification strategy and the credit policy strategy according to specific circumstances.

[0052] S30: According to the customer risk classification code and the order risk classification code, calling the decision engine for type matching to obtain the loan review type corresponding to the loan review information.

[0053] Exemplarily, the decision engine includes various strategies required for intelligent order distribution, such as customer risk classification strategy, credit policy strategy, loan review type strategy, order distribution strategy, etc. The decision engine can quantitatively classify the loan review information through corresponding matching of the loan review type strategy according to specific circumstances.

[0054] S40: According to the loan review type, calling the decision engine for order distribution matching to determine the loan review officer corresponding to the loan review information.

[0055] Exemplarily, the decision engine includes various strategies required for intelligent order distribution, such as customer risk classification strategy, credit policy strategy, loan review type strategy, order distribution strategy, etc. The decision engine can distribute the loan review information to the appropriate loan review officer through corresponding matching of the order distribution strategy according to specific circumstances of the loan review information and the loan review officer.

[0056] In this embodiment, the loan review order distribution method based on the decision engine automatically matches the most suitable loan review type and the decision-making optimal responsible loan review officer through multiple calls to the decision engine according to different dimensional attributes of the customer and the order, executes differentiated loan review, improves the efficiency of loan review, reduces the cost of loan review, and provides better services for customers. Optionally, the differentiated loan review refers to automatically matching the customer self-loan review, remote video loan review, loan review officer on-site loan review, and double on-site loan review types according to the order customer information and credit policy, and executing differentiated and refined credit audit process.

[0057] In an embodiment, the step of performing S20: calling the decision engine for strategy matching according to the customer basic information and the order information to correspondingly obtain a customer risk classification code and an order risk classification code in the loan review dispatch method based on the decision engine comprises:

[0058] S21: querying a customer credit record according to a customer ID number of the customer basic information;

[0059] S22: calling a customer risk classification strategy set in the decision engine, and performing rule matching of the risk strategy set according to the customer credit record to obtain a matched customer risk classification code.

[0060] Optionally, the customer credit record comprises at least one of a customer type, a region, external credit investigation, asset condition, internal credit investigation, internal historical purchase record, and historical repayment information.

[0061] In the embodiment, the loan review dispatch method based on the decision engine calls a customer risk classification strategy set in the decision engine according to a customer ID number to query external credit investigation, internal credit investigation, internal historical purchase record, and historical repayment information of the customer, and comprehensively calculates a customer risk classification by matching the customer information with rules in the strategy set to obtain a customer risk classification code.

[0062] In an embodiment, the step of performing S22: calling a customer risk classification strategy set in the decision engine, and performing rule matching of the risk strategy set according to the customer credit record to obtain a matched customer risk classification code in the loan review dispatch method based on the decision engine comprises:

[0063] S220: reading a historical customer credit record, and performing first weight assignment according to a first weight dimension of the historical customer, wherein the first weight dimension comprises a customer type, a region, external credit investigation, asset condition, internal credit investigation, internal historical purchase record, and historical repayment information;

[0064] S221: classifying different customers into a plurality of risk classifications according to the first weight assignment, and performing customer risk classification coding on the plurality of risk classifications;

[0065] S222: determining a risk classification code corresponding to the loan review information according to the first weight dimension of the customer credit record of the loan review information.

[0066] Exemplarily, the customer risk classification strategy set divides customers into five categories, i.e., A, B, C, D and E, each category including three sub-categories, i.e., 0, 1 and 2, totaling 15 customer risk classifications. Each customer risk classification is comprehensively evaluated from multiple dimensions, such as customer type, customer asset condition, enterprise internal credit investigation condition, customer historical repayment condition and the like, as the first weight dimension to determine the customer risk classification. Optionally, the customer risk classification of a government customer can be coded as A01, and the customer risk classification of a company alliance customer with normal enterprise internal credit investigation condition and no overdue historical purchase record can be coded as B01.

[0067] In an embodiment, the step of performing S20: calling the decision engine for strategy matching according to the customer basic information and the order information to correspondingly obtain a customer risk classification code and an order risk classification code includes:

[0068] S23: performing rule matching of the credit policy strategy set in the decision engine on the order information according to the purchase financial scheme of the order information to obtain a matched order risk classification code.

[0069] Optionally, the purchase financial scheme includes at least one of a purchase equipment type, an order amount, a down payment ratio, a financial scheme type and a guarantee condition.

[0070] In this embodiment, the loan review and order dispatch method based on the decision engine calls the credit policy strategy set in the decision engine according to the equipment information and the financial scheme, and comprehensively calculates the order risk classification by matching the order information with the rules in the strategy set to obtain the order risk classification code.

[0071] In an embodiment, the step of performing S23: calling the decision engine for strategy matching according to the customer basic information and the order information to correspondingly obtain a customer risk classification code and an order risk classification code includes:

[0072] S230: reading the purchase financial scheme of the historical customer, and respectively performing second weight assignment according to the second weight dimension of the historical customer, the second weight dimension including a purchase equipment type, an order amount, a down payment ratio, a financial scheme type and a guarantee condition;

[0073] S231: classifying different orders into multiple risk classifications according to the second weight assignment, and performing order risk classification coding on the multiple risk classifications;

[0074] S232: determining the order risk classification code corresponding to the order information according to the second weight dimension of the purchase financial scheme of the order information.

[0075] Optionally, the credit policy strategy set is comprehensively evaluated from multiple dimensions such as purchase equipment type, down payment ratio, financial scheme type, guarantee condition, etc. as the second weight dimension to finally determine a total of 21 order risk classifications. Exemplarily, the order risk classification of a government customer order can be coded as A, and the order risk classification of an enterprise customer purchasing a large excavator with a down payment of 50%, using a financial lease method, and with the actual controller of the enterprise guaranteeing can be coded as B.

[0076] In an embodiment, the loan review dispatching method based on the decision engine comprises the following steps:

[0077] S31: calling a scorecard model, inputting the customer risk classification code and the order risk classification code, so that the scorecard model outputs a score for the loan review information;

[0078] S32: reading the loan review information and the score, and using the loan review type strategy set in the decision engine to perform rule matching to determine the loan review type corresponding to the loan review information.

[0079] In this embodiment, the decision engine uses the calculated customer risk classification code and order risk classification code as variables for the scorecard model, comprehensively calculates the final score of the customer, and matches the rules in the loan review type strategy set to calculate the loan review type. Exemplarily, the scorecard model uses the customer risk classification and the order risk classification as two dimensions, and calculates the risk score of the customer and the order according to different weighting ratios of different business departments and different product types, which is used for judgment of the loan review type strategy. Optionally, the products produced by some business departments do not need to be refined to product types for judgment, and the products of different types produced by some business departments have large differences, so different weights are allocated according to the product types.

[0080] In an embodiment, the loan review type in the loan review dispatching method based on the decision engine comprises at least one of customer self-review, remote video loan review, loan review specialist on-site loan review, and double on-site loan review.

[0081] Exemplarily, if the loan review type is one of “customer self-review”, “remote video loan review”, “loan review specialist on-site loan review”, and “double on-site loan review”, then 1, 2, 3, and 4 are set up for four types of scores corresponding to four types of loan reviews. The loan review type strategy set is comprehensively evaluated from multiple dimensions such as customer risk classification, customer location, purchase amount, and purchase equipment type.

[0082] Exemplarily, for a new customer with a first purchase amount of 10 million yuan or above, a customer risk classification of B0, an order risk of E, and a loan review type score of 4, corresponding to a two-person on-site loan review, the first purchase amount is 1 million yuan or below, the purchase equipment type is an excavator, the customer risk classification is B0, the order risk is C, and the loan review type score is 2, corresponding to a remote video loan review. The first purchase amount is 20,000 yuan or below, the purchase equipment type is a tractor, the customer risk classification is B0, the order risk is B, and the loan review type score is 1, corresponding to a customer self loan review.

[0083] In an embodiment, the loan review dispatching method based on the decision engine comprises the following steps:

[0084] S41: calling a decision tree model, inputting the loan review information, and obtaining the credit manager qualification requirements corresponding to the loan review information;

[0085] S42: calling the current workloads of the credit managers meeting the credit manager qualification requirements, and determining the credit manager with the least current workload as the loan review specialist corresponding to the loan review information.

[0086] In this embodiment, the loan review dispatching method based on the decision engine matches the optimal loan review specialist through the decision tree model in the decision engine flow according to different requirements of different loan review types, through index variables such as the loan review specialist level and the service length.

[0087] Exemplarily, if the loan review type is a two-person loan review, the decision tree model matches all credit managers in the region of the order according to the senior credit manager level required by the two-person loan review, the assisting credit manager level of 2 years of service as an ordinary credit manager, and if multiple credit managers meeting the review conditions are matched, the credit managers are sorted in ascending order according to the workloads, and the credit manager with the least workload is determined as the loan review person in charge of the order.

[0088] As described above, the loan review dispatching method based on the decision engine can provide better loan review services for customers, better control customer risks, give ordinary loan review specialists the opportunity to participate in complex order loan reviews, accumulate loan review experience, and improve business capabilities faster by calling the decision engine multiple times, inputting the variables and parameters required by the decision engine, calling the internal strategy set of the decision engine, and calculating the optimal loan review scheme according to the rules and scorecard models.

[0089] Second embodiment

[0090] On the other hand, the application also provides a server.

[0091] In an embodiment, the server comprises a processor and a storage medium connected to each other. The storage medium is configured to store a computer program, and the processor is configured to read the computer program and run the computer program to enable the server to implement the decision engine-based loan review assignment method as described above.

[0092] Exemplarily, the intelligent assignment system used by the server in implementing the loan review assignment method comprises:

[0093] The intelligent assignment system automatically determines the applicable loan review type of an order according to the customer, order, and other dimension attributes implanted in the system, and matches the optimal responsible loan review officer according to the level, responsible area, and workload of the loan review officer.

[0094] The decision engine comprises various strategies required for intelligent assignment, such as customer risk classification strategy, credit policy strategy, loan review type strategy, and assignment strategy.

[0095] The intelligent assignment system described above calls the decision engine multiple times, inputs the variables and parameters required by the decision engine, and the decision engine calls the internal strategy set to calculate the optimal solution according to the rules and scorecard model and returns the corresponding interface code. The specific execution steps are as follows:

[0096] S1 The CRM system pushes the business opportunity information to the financial loan review platform through an interface, and the financial loan review platform transmits the customer basic information and order information to the intelligent assignment system (the intelligent assignment system sequentially calls the strategy set and model configured on the decision engine platform for automatic decision-making).

[0097] S2 After receiving the information from the financial loan review platform, the intelligent assignment system queries the customer external credit, enterprise internal credit, and enterprise internal historical purchase record information according to the customer ID, calls the customer risk classification strategy set in the decision engine, matches the customer information with the rules in the strategy set, and comprehensively calculates the customer risk classification and returns the customer risk classification code; the customer risk classification strategy set divides the customers into five categories, i.e., A, B, C, D, and E, each category includes three subcategories, i.e., 0, 1, and 2, and a total of 15 customer risk classifications. Each customer risk classification is comprehensively evaluated from multiple dimensions, such as customer type, customer asset situation, enterprise internal credit situation, and customer historical repayment situation, to finally determine the customer risk classification.

[0098] For example, a government customer is directly returned with the customer risk classification A01 code; for example, a company alliance customer with normal enterprise internal credit and no overdue historical purchase record is returned with the customer risk classification B01 code.

[0099] S3 intelligent dispatching system calls credit policy strategy set in decision engine according to equipment information, financial scheme and other indicators, matches order information with rules in the strategy set, comprehensively calculates order risk classification, and returns order risk classification code;

[0100] The credit policy strategy set comprehensively evaluates from multiple dimensions such as purchase equipment type, down payment ratio, financial scheme type, guarantee condition, and finally determines 21 order risk classifications in total from A to U.

[0101] For example, the order of a government customer is directly returned with order risk classification A code.

[0102] For example, an enterprise customer purchases a large excavator with a down payment of 50%, adopts a financial lease mode, and the actual controller of the enterprise guarantees, and the order risk classification B code is returned.

[0103] S4 decision engine takes the index results calculated by S2 and S3 as the input parameters of the scorecard model variable, comprehensively calculates the final score (1\2\3\4, four scores) of the customer, matches the rules in the loan review type strategy set, calculates the loan review type as one of "customer self-loan review", "remote video loan review", "loan review officer on-site loan review", and "two-person on-site loan review", and returns the loan review type code.

[0104] The scorecard model takes two dimensions of customer risk classification and order risk classification, adopts different weighting ratios according to different business departments and different product types (some products of some business departments do not need to be refined to product types for judgment, and some business departments have large differences between different types of products, so product types are still needed for judgment), and calculates the risk score of the customer and the order.

[0105] The loan review type strategy set comprehensively evaluates from multiple dimensions such as customer risk classification, customer location, purchase amount, and purchase equipment type.

[0106] At this time, the loan review type code can be returned.

[0107] For example, a new customer with a first purchase amount of more than 10 million yuan, a customer risk classification of B0, and an order risk of E returns a loan review type of 4, i.e., two-person on-site loan review.

[0108] For example, a new customer with a first purchase amount of less than 1 million yuan, a purchase equipment type of excavator, a customer risk classification of B0, and an order risk of C returns a loan review type of 2, i.e., remote video loan review.

[0109] For example, a new customer with a first purchase amount of less than 200,000 yuan, a purchase equipment type of tractor, a customer risk classification of B0, and an order risk of B returns a loan review type of 1, i.e., customer self-loan review.

[0110] S5 intelligent dispatching system matches the optimal loan review officer according to the different requirements of different loan review types judged by S4 through indicators such as loan review officer level and length of service, and finally returns the decision result to the upstream financial loan review platform. Alternatively, the score of the scorecard is used for the judgment of the loan review type strategy.

[0111] For example, the loan review type strategy set returns the loan review type as 4, double-person loan review, and the decision tree model matches all credit managers in the region where the order is located according to the requirement of the senior credit manager level of the main responsible credit manager for double-person loan review, and the level 2 ordinary credit manager within 2 years of employment (different credit manager level requirements), and if multiple credit managers meeting the review conditions are matched, the credit managers are sorted in ascending order according to the workload, and the credit manager with the least workload is taken as the loan review person in charge of the order.

[0112] Exemplarily, when the server implements the loan review dispatching method based on the decision engine, the steps of the video loan review include:

[0113] T1 The customer opens the video loan review applet, fills in the mobile phone number and video invitation code, clicks "call video officer", and the system calls the video officer through IM and instant call signaling;

[0114] T2 The video officer receives the IM and instant call signaling on the financial loan review platform PC end, automatically opens the video call answering page and the loan review work order page, and the video call page is floated on the loan review work order page, which can be freely placed by dragging, and the PC end plays a bell sound to remind the video officer to answer;

[0115] T3 The video officer clicks the answer button, creates a video call room for real-time audio and video service, and automatically pulls the customer applet end user and the video officer PC end user into the video call room;

[0116] T4 After the video call of both parties enters the room, the mixed flow transcoding and video recording service is started, the video and audio information of both parties is collected, the audio and video information of both parties is mixed and transcoded into one video, and the video stream is recorded into a video file;

[0117] T5 After the video call is successfully established, the video officer guides the customer to take photos of the front and back of the ID card, and the system automatically calls the OCR recognition service to identify the customer's name, ID number, and valid time;

[0118] T6 The video officer guides the customer to take a photo of the customer's face, and the system uses the customer's ID card photo and face photo to call the AI face recognition service for comparison to verify the customer's identity;

[0119] T7 If the customer does not bring an ID card, the customer provides an ID card number and name, the video officer guides the customer to take a customer face photo, calls an AI face recognition service, adopts a person and certificate verification method, compares the customer face photo with the ID card photo of the authoritative database, and verifies the customer;

[0120] T8 After the customer's identity is verified, the video officer communicates with the customer through video dialogue about the customer's purchase equipment information, financial scheme, customer assets, and intended use of the equipment, etc. During the conversation, the collected customer loan review information can be entered on the loan review work order page, and after the loan review information collection is completed, the customer or the officer hangs up the video call, and the video loan review is completed.

[0121] T9 The customer's ID card front and back photos, face photos, face comparison results, and double recording video files generated during the video call process are displayed on the loan review work order page, which facilitates subsequent review and restoration of the original loan review process, and is conducive to supporting possible subsequent legal litigation cases.

[0122] As described above, the server implements the loan review dispatching method based on the decision engine as described above, which creates four types of loan reviews: "customer self-loan review", "remote video loan review", "loan review specialist on-site loan review", and "two-person on-site loan review". Through the decision engine, various strategies required for intelligent dispatching are formulated, including customer risk classification strategy, credit policy strategy, loan review type strategy, and dispatching strategy. Through the intelligent dispatching system, the applicable loan review type of the order is automatically determined according to the customer and order dimension attributes implanted in the system, and the optimal responsible loan review specialist is matched according to the level, responsible area, and workload of the loan review specialist. Based on video streaming media and AI face recognition technology, a remote video loan review technical solution is realized, and the remote video loan review and loan review work order are creatively combined, realizing the synchronization of remote video loan review and loan review report writing, and displaying the customer's ID card front and back photos, face photos, face comparison results, and double recording video files generated during the video call process on the loan review work order page, which facilitates subsequent review and restoration of the original loan review process.

[0123] Third embodiment

[0124] On the other hand, the application also provides a storage medium.

[0125] In an embodiment, the storage medium stores a computer program, and the computer program is executed by a processor to implement the steps of the loan review dispatching method based on the decision engine as described above.

[0126] As described above, the decision engine-based loan review and assignment method, server and storage medium provided by the present application are applied to actual projects of some financing leasing companies. Compared with the traditional loan review method of loan review officers on site, remote video loan review can save the travel cost of loan review officers going to the customer site, and the overall loan review cost can be reduced by more than 98%. Compared with the traditional loan review method of loan review officers on site, the loan review officers do not need to go to the customer site, and can complete the loan review task without going out, and the overall loan review time can be reduced by more than 90%. For some large orders and high-risk customers, compared with the loan review method of loan review officers on site, double loan review can provide better loan review services for customers, better control customer risks, and at the same time give ordinary loan review officers the opportunity to participate in complex order loan review, accumulate loan review experience, and improve business ability faster.

[0127] It should be noted that in the present application, step codes such as S10, S20, etc. are used, the purpose of which is to more clearly and briefly describe the corresponding content, and does not constitute a substantial limitation on the order. Those skilled in the art may, for example, first perform S20 and then perform S10, etc. when implementing, but these should be within the scope of protection of the present application.

[0128] In the embodiments of the server and storage medium provided by the present application, any of the above method embodiments can contain all the technical features, and the description and explanation content is basically the same as that of the above method embodiments, which will not be repeated here.

[0129] The embodiments of the present application also provide a computer program product, which includes computer program code, when the computer program code runs on a computer, so that the computer executes the method in various possible embodiments as above.

[0130] The embodiments of the present application also provide a chip, which includes a memory and a processor, the memory is used to store a computer program, and the processor is used to call and run the computer program from the memory, so that the device installed with the chip executes the method in various possible embodiments as above.

[0131] It can be understood that the above scenarios are only examples and do not constitute a limitation on the application scenarios of the technical solutions provided by the embodiments of the present application. The technical solutions provided by the embodiments of the present application are also applicable to other scenarios. For example, those skilled in the art can know that with the evolution of system architecture and the appearance of new business scenarios, the technical solutions provided by the embodiments of the present application are also applicable to similar technical problems.

[0132] The above sequence numbers of the embodiments of the present application are only for description, and do not represent the advantages and disadvantages of the embodiments.

[0133] The steps in the method of the embodiments of the present application can be adjusted, combined and reduced in sequence according to actual needs.

[0134] The units in the device of the embodiments of the present application can be combined, divided, and deleted according to actual needs.

[0135] In the present application, for the same or similar term concept, technical solution and / or application scenario description, generally only the first time is described in detail, and for the sake of brevity, the repeated description is generally not repeated, and for the understanding of the technical solutions of the present application, the same or similar term concept, technical solution and / or application scenario description which is not described in detail can be referred to the previous related description.

[0136] In the present application, the description of each embodiment has its own emphasis, and the part not described or recorded in a certain embodiment can be referred to the related description of other embodiments.

[0137] The technical features of the technical solutions of the present application can be combined arbitrarily, in order to make the description simple, the possible combinations of each technical feature in the above embodiments are not described, however, as long as the combination of these technical features does not exist contradictory, it should be considered as the range recorded in the present application.

[0138] The above is only the preferred embodiment of the present application, and does not limit the patent scope of the present application, any equivalent structure or equivalent process transformation using the content of the present application specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present application.

Claims

1. A method for loan review and dispatching based on a decision engine, characterized in that, The method comprises the following steps: in response to obtaining the loan review information, extracting the customer basic information and order information in the loan review information; according to the customer basic information and the order information, calling a decision engine for strategy matching to correspondingly obtain a customer risk classification code and an order risk classification code; according to the customer risk classification code and the order risk classification code, calling the decision engine for type matching to obtain a loan review type corresponding to the loan review information; according to the loan review type, calling the decision engine for order matching to determine a loan review officer corresponding to the loan review information; the step of calling the decision engine for type matching according to the customer risk classification code and the order risk classification code to obtain the loan review type corresponding to the loan review information comprises: calling a scorecard model, inputting the customer risk classification code and the order risk classification code, so that the scorecard model outputs a score for the loan review information; reading the loan review information and the score, and using a loan review type strategy set in the decision engine for rule matching to determine the loan review type corresponding to the loan review information; the loan review type comprises at least one of customer self-loan review, remote video loan review, loan review officer on-site loan review and double on-site loan review.

2. The decision engine based loan review assignment method of claim 1, wherein, the step of calling the decision engine for strategy matching according to the customer basic information and the order information to correspondingly obtain the customer risk classification code and the order risk classification code comprises: querying a customer credit record according to a customer ID number in the customer basic information; calling a customer risk classification strategy set in the decision engine, and performing rule matching of a risk strategy set according to the customer credit record to obtain a matched customer risk classification code; wherein the customer credit record comprises at least one of customer type, region, external credit, asset situation, internal credit of enterprise, historical procurement record and historical repayment information of enterprise. 3.The decision engine based loan review assignment method of claim 2, wherein, the step of calling the decision engine for strategy matching according to the customer basic information and the order information to correspondingly obtain the customer risk classification code and the order risk classification code comprises: reading a historical customer credit record, and performing first weight assignment according to a first weight dimension of the historical customer, the first weight dimension comprising customer type, region, external credit, asset situation, internal credit of enterprise, historical procurement record and historical repayment information of enterprise; classifying different customers into a plurality of risk classifications according to the first weight assignment, and performing customer risk classification coding on the plurality of risk classifications; determining the customer risk classification code corresponding to the loan review information according to the first weight dimension of the customer credit record of the loan review information.

4. The decision engine based loan review assignment method of claim 1, wherein, the step of calling the decision engine for strategy matching according to the customer basic information and the order information to correspondingly obtain the customer risk classification code and the order risk classification code comprises: according to a procurement financial scheme of the order information, using a credit policy strategy set in the decision engine to perform rule matching of a credit strategy set on the order information to obtain a matched order risk classification code; The procurement financial scheme includes at least one of a procurement equipment type, an order amount, a down payment ratio, a financial scheme type, and a guarantee condition.

5. The decision engine based loan review assignment method of claim 4, wherein, The step of using the credit policy strategy set in the decision engine to perform rule matching on the order information to obtain a matched order risk classification code includes: Reading the procurement financial scheme of the historical customer, and performing second weight assignment on the historical customer according to a second weight dimension, which includes a procurement equipment type, an order amount, a down payment ratio, a financial scheme type, and a guarantee condition; According to the second weight assignment, different orders are classified into a plurality of risk classifications, and the plurality of risk classifications are coded into order risk classification codes; According to the second weight dimension of the order information of the procurement financial scheme, the order risk classification code corresponding to the order information is determined.

6. The decision engine based loan review assignment method according to any one of claims 1-5, wherein, The step of calling the decision engine to perform order matching according to the loan review type to determine the loan review officer corresponding to the loan review information includes: Calling a decision tree model, inputting the loan review information, and obtaining the credit manager qualification requirement corresponding to the loan review information; Call the current workload of the credit manager who meets the credit manager qualification requirement, and determine the credit manager with the least current workload as the loan review officer corresponding to the loan review information.

7. A server, characterized by The server includes a processor and a storage medium connected to each other, wherein: The storage medium is used to store a computer program; The processor is used to read and run the computer program, so that the server implements the loan review order dispatching method based on the decision engine according to any one of claims 1-6.

8. A storage medium, in particular, the storage medium stores a computer program, and the computer program is executed by a processor to implement the steps of the loan review order dispatching method based on the decision engine according to any one of claims 1-6.

Images