A method, device, medium, and product for querying business indicators.
By generating, validating, and executing business execution scripts in a secure sandbox, the problems of numerical illusion and unauditability in financial indicator calculations using large language models are resolved, ensuring the accuracy and compliance of the calculation results.
Patent Information
- Application Number
- CN202611091760.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-07-22
- Publication Date
- 2026-11-17
AI Technical Summary
When large language models are used to calculate corporate financial indicators in financial risk control scenarios, numerical illusions are easily generated. The calculation process is not traceable and auditable and lacks an automatic error correction mechanism, resulting in calculation results that do not meet financial compliance requirements.
The first intelligent agent generates the business execution script, the second intelligent agent verifies the script, and the script is executed in a security sandbox to ensure the compliance and security of the script, and output auditable business operation results.
It achieves closed-loop verification and iterative optimization of business execution scripts, eliminates script data and logical deviations, ensures the accuracy and security of calculation results, and complies with financial compliance requirements.
Smart Images

Figure CN123118236A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the fields of financial technology and artificial intelligence, and in particular to a method, device, medium and product for querying business indicators. Background Technology
[0002] In financial risk control scenarios, the calculation of corporate financial indicators is a core task in credit assessment. The calculation of indicators such as current ratio and quick ratio must strictly follow unified accounting standards and bank risk control formulas. To reduce manual calculation costs and complete corporate risk assessments in batches, the industry has widely introduced large language models to handle financial indicator query services, relying on artificial intelligence to automatically analyze financial statement data, thereby improving the overall efficiency of risk control review.
[0003] Current conventional automation solutions rely solely on a single large language model to complete the entire process. After receiving data, the system autonomously performs data reading, formula derivation, and numerical calculations, directly outputting the final indicator value. Large models, based on probability-based text generation, are prone to mathematical illusions, and financial numerical calculations are susceptible to errors. The lack of standardized calculation scripts means the computational logic is a black box and cannot be audited, failing to meet financial compliance requirements. Furthermore, the absence of an independent verification and error correction loop means that errors in data retrieval and formula logic cannot be automatically identified and corrected, potentially misleading risk control decisions. Summary of the Invention
[0004] This invention provides a business indicator query method, device, medium, and product, which solves the problems of numerical illusion, lack of auditability of the calculation process, and lack of automatic error correction mechanism when large models are used to calculate enterprise financial indicators.
[0005] According to one aspect of the present invention, a business indicator query method is provided, the method comprising: The first intelligent agent receives query requests and business data, and generates a business execution script based on the query requests and business data. The query requests include business metrics. The business execution script is validated based on the second intelligent agent; When the verification passes, the business execution script is run through the preset security sandbox and the business calculation results are output.
[0006] Optionally, a business execution script is generated based on the query request and business data, including: retrieving business operation information corresponding to the business operation information from a preset business knowledge base based on business indicators; generating a first prompt word based on the business operation information; and generating a business execution script corresponding to the query request and business data through a first intelligent agent based on the first prompt word.
[0007] The advantages of this setup are: by standardizing business operation logic through a knowledge base, and by automatically generating business execution scripts through intelligent agents, the script generation process is simplified, human error is avoided, and reliable script support is provided for subsequent business processing and risk assessment.
[0008] Optionally, the business execution script is validated based on the second intelligent agent, including: generating a second prompt word, and using the second intelligent agent to validate the target dimension of the business execution script according to the second prompt word, and obtaining the validation result; wherein, the target dimension includes field matching validation, period consistency validation, logical correctness validation, hierarchical accuracy validation, and unit normalization validation.
[0009] The advantage of this setup is that it allows for a comprehensive check of the compliance and accuracy of business execution scripts, eliminating deviations in script data and logic.
[0010] Optionally, the method also includes: when the verification fails, generating verification exception information and sending the verification exception information to the first intelligent agent; and having the first intelligent agent regenerate the business execution script based on the verification exception information.
[0011] The advantage of this setup is that it enables closed-loop verification and iterative optimization of the business execution script, correcting any defects or problems in the script.
[0012] Optionally, the business execution script can be run through a preset security sandbox and the business calculation results can be output, including: running the business execution script in an isolated environment through the preset security sandbox, and outputting the corresponding business calculation results when the script runs successfully.
[0013] The advantage of this setup is that it enables the script to run in isolation and securely, avoiding the security risks associated with script execution.
[0014] Optionally, after running the business execution script through a preset security sandbox, the method further includes: generating error information when the script fails to run, and sending the error information to the first intelligent agent; regenerating the business execution script based on the error information by the first intelligent agent, until the verification passes and the script runs successfully, and then outputting the corresponding business calculation result.
[0015] The advantages of this setup are: it establishes a closed-loop iterative mechanism for script execution, fixes script execution anomalies, and ensures the stability and reliability of business operation processes.
[0016] Optionally, before receiving the query request and business data based on the first intelligent agent, the method further includes: obtaining an unstructured business document; generating a third prompt word; using the third intelligent agent to perform text recognition and layout analysis on the unstructured business document based on the third prompt word to obtain analysis data; and converting the analysis data into structured business data.
[0017] The advantage of this setup is that it enables the standardization of unstructured business data and unifies the format of business data.
[0018] 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; The memory stores a computer program that can be executed by the at least one processor, which is then executed by the at least one processor to enable the at least one processor to execute a business indicator query method according to any embodiment of the present invention.
[0019] According to another aspect of the present invention, a computer storage medium is provided, the computer storage medium storing computer instructions, the computer instructions being used to cause a processor to execute and implement a business indicator query method as described in any embodiment of the present invention.
[0020] According to another aspect of the present invention, a computer program product is provided, the computer program product comprising a computer program that, when executed by a processor, implements a business indicator query method according to any embodiment of the present invention.
[0021] The technical solution of this invention transforms query requests carrying business indicators and business data into business execution scripts through a first intelligent agent. This separates numerical calculations from the large model, avoiding mathematical illusions from the source and preserving standardized calculation logic for easy traceability. A second intelligent agent separately verifies the business execution scripts, proactively checking field matching and formula definitions to ensure compliance with risk control standards and promptly intercepting logical errors. A security sandbox runs only verified scripts, isolating execution to guarantee data security and relying on code determinism to output accurate and auditable business calculation results.
[0022] 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
[0023] 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.
[0024] Figure 1This is a flowchart of a business indicator query method provided in Embodiment 1 of the present invention; Figure 2 This is a flowchart of another business indicator query method provided in Embodiment 2 of the present invention; Figure 3 This is a schematic diagram of the structure of a business indicator query device provided in Embodiment 3 of the present invention; Figure 4 This is a schematic diagram of the structure of an electronic device that implements a business indicator query method according to an embodiment of the present invention. Detailed Implementation
[0025] 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 of the present invention. 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.
[0026] 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.
[0027] Example 1 Figure 1 This is a flowchart illustrating a business indicator query method provided in Embodiment 1 of the present invention. This embodiment is applicable to scenarios involving automated query and calculation of enterprise financial indicators. The method can be executed by a business indicator query device, which can be implemented in hardware and / or software and can be configured within a computer controller. Figure 1 As shown, the method includes: S110. The first intelligent agent receives query requests and business data, and generates a business execution script based on the query requests and business data, wherein the query requests include business metrics.
[0028] The first intelligent agent is a large-scale intelligent agent used to convert natural language query results into executable code scripts. Instead of directly outputting results, the first intelligent agent generates execution scripts based on user query requests and structured business data, combined with an external knowledge base. Its role is to eliminate the numerical calculation work of the large model, removing the mathematical illusion at its source, and relying on a code interpreter to ensure computational accuracy. A query request refers to a user's query instruction regarding business indicators, expressed in natural language or structured form, such as "analyze the quick ratio in 2025." The query request can include the business indicator to be calculated, the time range, and the data dimensions. The business data is the original financial information of the enterprise to be analyzed, which is preprocessed by the data structuring intelligent agent to obtain structured data. The execution script is standardized code generated by the first intelligent agent and can be executed by the program; it is the carrier of indicator calculations. Business indicators are quantitative standards for measuring business performance. In the financial field, these may be current ratios and quick ratios, while in other fields they may include revenue growth rate and user retention rate. Each indicator has a clear calculation formula and data source, and must be injected into the system through an external knowledge base to ensure computational compliance.
[0029] Optionally, a business execution script is generated based on the query request and business data, including: retrieving business operation information corresponding to the business operation information from a preset business knowledge base based on business indicators; generating a first prompt word based on the business operation information; and generating a business execution script corresponding to the query request and business data through a first intelligent agent based on the first prompt word.
[0030] The pre-defined business knowledge base is an external knowledge base that stores specific business rules, indicator calculation formulas, etc. For example, the quick ratio calculation formula is: Quick Ratio = (Current Assets - Inventory) / Current Liabilities. Business calculation information is the specific calculation formulas and related rules corresponding to the business indicators retrieved from the knowledge base. The first prompt is an instruction text that integrates business calculation information, user query intent, and business data context, used to guide the first intelligent agent in generating code.
[0031] Specifically, when a user enters a query request, such as "Please analyze the company's quick ratio in 2025," the system first extracts the business metric, namely the quick ratio, from the request. Then, it searches a pre-defined business knowledge base to obtain the corresponding business calculation information, i.e., the formula for calculating the quick ratio. The system integrates this business calculation information with the specific year 2025 from the user's query and the source of the business data to generate a first prompt. This first prompt explicitly instructs the code analyst agent to calculate the quick ratio for the specified year based on the given formula and data. Finally, the first agent, based on the generated first prompt and the input structured business data, generates the query request and the corresponding business execution script.
[0032] S120. Verify the business execution script based on the second intelligent agent.
[0033] The second agent is an independent code logic verification module that acts as a code reviewer. It does not execute code but only checks the logical correctness of the business execution script through static analysis. Its role is to build an adversarial verification loop of "generation-audit-correction" to avoid calculation deviations caused by code logic errors.
[0034] Optionally, the business execution script is validated based on the second intelligent agent, including: generating a second prompt word, and using the second intelligent agent to validate the target dimension of the business execution script according to the second prompt word, and obtaining the validation result; wherein, the target dimension includes field matching validation, period consistency validation, logical correctness validation, hierarchical accuracy validation, and unit normalization validation.
[0035] The second prompt is the instruction text given to the second intelligent agent, clearly specifying the verification task and the target dimensions to be checked. The target dimensions specifically include field matching verification, period consistency verification, logical correctness verification, hierarchical accuracy verification, and unit normalization verification. Field matching verification checks whether the data column names referenced in the code match the account names in the original financial documents. Period consistency verification ensures that the time period involved in the code matches the time range of the user query or the original data. Logical correctness verification verifies whether the calculation formulas in the code conform to specific business rules. Hierarchical accuracy verification confirms whether the code accurately references hierarchical financial statement data. Unit normalization verification checks whether the code has uniformly processed the data units in the financial statements.
[0036] Specifically, the system generates a second prompt word containing all verification dimension standards. This second prompt word guides the second agent to perform multi-dimensional verification of the business execution script, clearly listing the specific requirements for the five verification dimensions: field matching, period consistency, logical correctness, hierarchical accuracy, and unit normalization. The system then sends the second prompt word and the business execution script to the second agent. The second agent performs verification according to the requirements in the second prompt word. In the field matching verification stage, it compares the field names in the code with the account names in the original financial statements. In the period consistency verification stage, it checks whether the time fields referenced in the code, such as "2025," match the time requirements in the user's query request. In the logical correctness verification stage, it checks whether the calculation formulas in the code, such as the quick ratio calculation, conform to the preset business rules. In the hierarchical accuracy verification stage, it confirms whether the summary fields referenced in the code correspond to the correct summary hierarchy in the financial statements. In the unit normalization verification stage, it checks whether there is any data with different units in the code. After the second agent completes the validation across all dimensions, it generates validation results. If it finds any issues, such as field mismatches or incorrect formulas, it reports the specific problems to the first agent, which then corrects the business execution script based on the feedback. Once all validations pass, the script execution phase begins.
[0037] Optionally, the method also includes: when the verification fails, generating verification exception information and sending the verification exception information to the first intelligent agent; and having the first intelligent agent regenerate the business execution script based on the verification exception information.
[0038] Specifically, when the second agent detects these problems that cause validation to fail, it generates corresponding validation exception information, such as "Field name mismatch, please correct," and then sends the validation exception information to the first agent. After receiving the validation exception information, the first agent triggers a self-reflection mechanism to analyze the cause of the error. If it determines that the field name is incorrect, it will readjust the field references in the code according to the exception information and regenerate a business execution script that meets the requirements to ensure the consistency of the code logic with the data source and accounting standards.
[0039] S130. When the verification passes, the business execution script is run through the preset security sandbox and the business calculation results are output.
[0040] The pre-defined security sandbox is an independent, isolated code execution environment that only allows validated scripts to run. It isolates scripts from core system business data, mitigating the risk of malicious code and capturing runtime error messages to support the system's automatic error correction loop. The business calculation result refers to the standardized metric values output by the security sandbox after executing the script, accompanied by a complete explanation of the code calculation process.
[0041] Optionally, the business execution script can be run through a preset security sandbox and the business calculation results can be output, including: running the business execution script in an isolated environment through the preset security sandbox, and outputting the corresponding business calculation results when the script runs successfully.
[0042] Specifically, after the business execution script evades auditing through static code manipulation by the second agent, it is sent to a pre-defined security sandbox. The security sandbox runs the script in an isolated environment, ensuring that script execution does not interfere with other parts of the system and preventing the script from posing risks to core system data or the external environment. When the script runs successfully, the security sandbox outputs the corresponding business calculation results, such as the specific quick ratio value. Simultaneously, the system also outputs an explanation of the calculation process, translating the code logic into natural language, thus white-boxing the calculation process and facilitating auditing and verification.
[0043] Optionally, after running the business execution script through a preset security sandbox, the method further includes: generating error information when the script fails to run, and sending the error information to the first intelligent agent; regenerating the business execution script based on the error information by the first intelligent agent, until the verification passes and the script runs successfully, and then outputting the corresponding business calculation result.
[0044] The error message is an error log generated by the security sandbox when a script fails to run, containing information such as the error type and location.
[0045] Specifically, when a business execution script fails to run in the security sandbox, such as due to a data field reference error leading to a key error or a division by zero error, the security sandbox generates an error message containing information such as the error type and location, and then sends the error message to the first agent. Upon receiving the error message, the first agent triggers a self-reflection mechanism to analyze the reason for the code failure. If it determines that the error is due to a data field reference error, it will readjust the field references or formula logic in the code based on the error message and regenerate the corrected business execution script. Then, the newly generated business execution script is sent to the second agent for verification. If the verification passes, it is then sent to the security sandbox for execution. If it still fails, the iterative process of generation, auditing, and correction is triggered again. This process continues until the newly generated script passes the logic verification of the auditing critic agent and can run successfully in the security sandbox, finally outputting the corresponding business calculation result.
[0046] The technical solution of this invention transforms query requests carrying business indicators and business data into business execution scripts through a first intelligent agent. This separates numerical calculations from the large model, avoiding mathematical illusions from the source and preserving standardized calculation logic for easy traceability. A second intelligent agent separately verifies the business execution scripts, proactively checking field matching and formula definitions to ensure compliance with risk control standards and promptly intercepting logical errors. A security sandbox runs only verified scripts, isolating execution to guarantee data security and relying on code determinism to output accurate and auditable business calculation results.
[0047] Example 2 Figure 2 This is a flowchart of a business indicator query method provided in Embodiment 2 of the present invention. This embodiment adds a business data conversion process based on Embodiment 1. The specific content of steps S230-S250 is largely the same as steps S110-S130 in Embodiment 1, and therefore will not be repeated in this embodiment. Figure 2 As shown, the method includes: S210. Obtain unstructured business documents.
[0048] Unstructured business documents refer to original files of corporate financial statements and report images that have not undergone standardization processing.
[0049] S220. Generate a third prompt word. Through a third intelligent agent, perform text recognition and layout analysis on unstructured business documents based on the third prompt word to obtain analysis data, and then convert the analysis data into structured business data.
[0050] The third intelligent agent is a data-structured intelligent agent, an independent agent used to process raw financial statement documents. The third prompt word is a customized instruction text for text recognition and layout analysis tasks, used to constrain the third intelligent agent to fully preserve the hierarchy of financial tables and distinguish different report items. Text recognition can be optical character recognition technology, responsible for extracting all text and numerical content within the document. Layout analysis involves recognizing and organizing the table positions, row and column levels, and report categories within the document. Analyzed data refers to the intermediate data of the raw text tables with position and hierarchy information generated after text recognition and layout analysis.
[0051] Specifically, the third prompt explicitly requires distinguishing between the balance sheet and the income statement, retaining constraints such as account names, annual values, and table hierarchy. The system can input the third prompt along with unstructured business documents into a third-party AI agent. The agent uses text recognition technology to extract all financial figures and account text from the document, while simultaneously using layout analysis technology to locate table areas, distinguishing table headers, row items, and annual columns to avoid data confusion between different reports. After recognition and parsing, analytical data with layout hierarchy information is generated. Then, the third-party AI agent performs a unified format conversion on the analytical data, cleaning up messy layouts and standardizing financial account names, transforming scattered text tables into standardized and readable structured business data. This structuring process solves the problem of original financial statements not being directly readable by code, providing a compliant and complete standardized data foundation for subsequent accurate calculation of financial indicators. It avoids errors in subsequent code reading due to messy original document layout and text recognition errors, forming a complete data error correction front-end defense in conjunction with the back-end audit and verification process.
[0052] S230. The first intelligent agent receives query requests and business data, and generates a business execution script based on the query requests and business data, wherein the query requests include business metrics.
[0053] Optionally, a business execution script is generated based on the query request and business data, including: retrieving business operation information corresponding to the business operation information from a preset business knowledge base based on business indicators; generating a first prompt word based on the business operation information; and generating a business execution script corresponding to the query request and business data through a first intelligent agent based on the first prompt word.
[0054] S240. Verify the business execution script based on the second intelligent agent.
[0055] Optionally, the business execution script is validated based on the second intelligent agent, including: generating a second prompt word, and using the second intelligent agent to validate the target dimension of the business execution script according to the second prompt word, and obtaining the validation result; wherein, the target dimension includes field matching validation, period consistency validation, logical correctness validation, hierarchical accuracy validation, and unit normalization validation.
[0056] Optionally, the method also includes: when the verification fails, generating verification exception information and sending the verification exception information to the first intelligent agent; and having the first intelligent agent regenerate the business execution script based on the verification exception information.
[0057] S250. When the verification passes, the business execution script is run through the preset security sandbox and the business calculation results are output.
[0058] Optionally, the business execution script can be run through a preset security sandbox and the business calculation results can be output, including: running the business execution script in an isolated environment through the preset security sandbox, and outputting the corresponding business calculation results when the script runs successfully.
[0059] Optionally, after running the business execution script through a preset security sandbox, the method further includes: generating error information when the script fails to run, and sending the error information to the first intelligent agent; regenerating the business execution script based on the error information by the first intelligent agent, until the verification passes and the script runs successfully, and then outputting the corresponding business calculation result.
[0060] The technical solution of this invention can completely retain all information from an enterprise's original financial statements by acquiring unstructured business documents, thus avoiding the loss of original data. By generating third prompt words adapted to the parsing task, the processing logic of the third intelligent agent can be standardized, constraining recognition and parsing standards. The third intelligent agent completes text recognition and layout analysis based on the prompt words, accurately extracting financial text and values from the document and distinguishing table levels, producing well-organized analytical data. By converting the analytical data into structured business data, the data format can be unified, allowing the data to be directly read and called by the subsequent first intelligent agent, eliminating the obstacle of the original messy document being unable to be directly processed.
[0061] Example 3 Figure 3 This is a schematic diagram of a business indicator query device provided in Embodiment 3 of the present invention. Figure 3 As shown, the device includes: a business execution script generation module 310, which is used to receive query requests and business data based on a first intelligent agent, and generate a business execution script based on the query requests and business data, wherein the query requests include business indicators; The script verification module 320 is used to verify the business execution script based on the second intelligent agent; The calculation result output module 330 is used to run the business execution script through a preset security sandbox and output the business calculation result when the verification is passed.
[0062] Optionally, the business execution script generation module 310 is specifically used for: retrieving business operation information corresponding to the business operation information from a preset business knowledge base based on business indicators; generating a first prompt word based on the business operation information; and generating a business execution script corresponding to the query request and business data through a first intelligent agent based on the first prompt word.
[0063] Optionally, the script verification module 320 is specifically used to: generate a second prompt word, and through a second intelligent agent, verify the target dimension of the business execution script based on the second prompt word to obtain the verification result; wherein, the target dimension includes field matching verification, period consistency verification, logical correctness verification, hierarchical accuracy verification, and unit normalization verification.
[0064] Optionally, the device further includes: a verification failure module, specifically used for: generating verification exception information when verification fails, and sending the verification exception information to the first intelligent agent; and regenerating the business execution script by the first intelligent agent based on the verification exception information.
[0065] Optionally, the calculation result output module 330 is specifically used to: run the business execution script in an isolated environment through a preset security sandbox, and output the corresponding business calculation result when the script runs successfully.
[0066] Optionally, the device further includes: a failure execution module, used to generate error information and send the error information to the first intelligent agent when the script fails to run after running the business execution script through a preset security sandbox; the first intelligent agent regenerates the business execution script according to the error information until the verification is passed and the running is successful, and then outputs the corresponding business calculation result.
[0067] Optionally, the device further includes: a business data conversion module, used to obtain unstructured business documents before receiving query requests and business data based on the first intelligent agent; generate a third prompt word; and through the third intelligent agent, perform text recognition and layout analysis on the unstructured business documents based on the third prompt word to obtain analysis data, and convert the analysis data into structured business data.
[0068] The technical solution of this invention transforms query requests carrying business indicators and business data into business execution scripts through a first intelligent agent. This separates numerical calculations from the large model, avoiding mathematical illusions from the source and preserving standardized calculation logic for easy traceability. A second intelligent agent separately verifies the business execution scripts, proactively checking field matching and formula definitions to ensure compliance with risk control standards and promptly intercepting logical errors. A security sandbox runs only verified scripts, isolating execution to guarantee data security and relying on code determinism to output accurate and auditable business calculation results.
[0069] The business indicator query device provided in this embodiment of the invention can execute a business indicator query method provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects of the execution method.
[0070] Example 4 Figure 4A schematic diagram of an electronic device 10 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 assistants, 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.
[0071] like Figure 4 As shown, the electronic device 10 includes at least one processor 11 and a memory, such as a read-only memory (ROM) or random access memory (RAM), communicatively connected to the at least one processor 11. The memory stores computer programs executable by the at least one processor. The processor 11 can perform various appropriate actions and processes based on the computer program stored in the ROM 12 or loaded into the RAM 13 from the storage unit 18. The RAM 13 can also store various programs and data required for the operation of the electronic device 10. The processor 11, ROM 12, and RAM 13 are interconnected via a bus 14. Input / output (I / O) interfaces are also connected to the bus 14.
[0072] Multiple components in electronic device 10 are connected to I / O interface 15, including: input unit 16, such as keyboard, mouse, etc.; output unit 17, such as various types of displays, speakers, etc.; storage unit 18, such as disk, optical disk, etc.; and communication unit 19, such as network card, modem, wireless transceiver, etc. Communication unit 19 allows electronic device 10 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.
[0073] Processor 11 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 11 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, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. Processor 11 performs the various methods and processes described above, such as a business metric query method.
[0074] In some embodiments, a business metric query method may be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 18. In some embodiments, part or all of the computer program may be loaded and / or installed on electronic device 10 via ROM 12 and / or communication unit 19. When the computer program is loaded into RAM 13 and executed by processor 11, one or more steps of the business metric query method described above may be performed. Alternatively, in other embodiments, processor 11 may be configured to execute a business metric query method by any other suitable means (e.g., by means of firmware).
[0075] 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.
[0076] 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.
[0077] 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.
[0078] 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).
[0079] 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.
[0080] 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.
[0081] 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.
[0082] 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 service indicator inquiry method, characterized by, include: The first intelligent agent receives query requests and business data, and generates a business execution script based on the query requests and the business data, wherein the query requests include business metrics; The second intelligent agent is used to verify the business execution script; When the verification passes, the business execution script is run through a preset security sandbox and the business calculation results are output.
2. The method of claim 1, wherein, The step of generating a business execution script based on the query request and the business data includes: Based on the business indicators, a search is performed in a preset business knowledge base to obtain business operation information corresponding to the business indicators; A first prompt word is generated based on the aforementioned business operation information; The first intelligent agent generates a business execution script corresponding to the query request and the business data based on the first prompt word.
3. The method of claim 1, wherein, The verification of the service execution script based on the second intelligent agent includes: A second prompt word is generated, and the second intelligent agent performs a verification of the target dimension of the business execution script based on the second prompt word to obtain the verification result; The target dimensions include field matching verification, period consistency verification, logical correctness verification, hierarchical accuracy verification, and unit normalization verification.
4. The method of claim 3, wherein, The method further includes: When the verification fails, a verification error message is generated and sent to the first intelligent agent; The first intelligent agent regenerates the business execution script based on the verification anomaly information.
5. The method of claim 1, wherein, The process of running the business execution script through a preset security sandbox and outputting the business calculation results includes: The business execution script is run in an isolated environment through a preset security sandbox. When the script runs successfully, the corresponding business calculation result is output.
6. The method of claim 5, wherein, After running the business execution script through a preset security sandbox, the method further includes: When the script fails to run, an error message is generated and sent to the first intelligent agent; The first intelligent agent regenerates the business execution script based on the error information until the verification is passed and the operation is successful, at which point the corresponding business calculation result is output.
7. The method according to any one of claims 1 to 6, characterized in that, Before the first intelligent agent receives the query request and business data, the method further includes: Obtain unstructured business documents; A third prompt word is generated. A third intelligent agent performs text recognition and layout analysis on the unstructured business document based on the third prompt word to obtain analysis data, and then converts the analysis data into structured business data.
8. An electronic device, comprising: The electronic device includes: At least one processor; and a memory communicatively connected to the at least one processor; 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 method of any one of claims 1-7.
9. A computer storage medium, characterized in that The computer storage medium stores computer instructions that are used to cause a processor to execute the method of any one of claims 1-7.
10. A computer program product, characterised in that, The computer program product includes a computer program that, when executed by a processor, implements the method according to any one of claims 1-7.
