With the advent of the digital age, traditional project management methods for high-tech projects are no longer sufficient to meet the needs of high-tech enterprises' management. The "National Informatization Development Strategic Outline" in the section "Focusing on Enhancing the Informatization Level of the Economy and Society" particularly emphasizes accelerating the integration and innovation of information technology with manufacturing technology, products, equipment, and comprehensively enhancing the intelligence level of enterprises' research and development, production, management, and services. Therefore, by leveraging information technology, from the perspective of the entire cycle of high-tech project research and development projects, a shared information platform is constructed to achieve the full-cycle management of project elements, thereby creating a high-level of informatization in enterprise research and development management and services, and promoting the high-quality development of high-tech enterprises' research and development activities.

Since the finance department took the lead in organizing the application, re-evaluation and enjoying the annual additional deduction policy for high-tech enterprises, there has been no unified management software system for the project establishment, implementation, time cost allocation and budget of high-tech research projects. A large amount of work relies on urging the personnel of each department to make modifications and reviews, which is time-consuming and labor-intensive. With the increasingly strict management requirements from the tax bureau and the science and technology bureau, an integrated information software for R&D management and cost management is urgently needed to improve work efficiency and quality.

Through the implementation of this functional module, in addition to completing the final research and management application of R&D management, a deep fundamental research is conducted on the secondary development capabilities and development achievements of the pan-micro platform. The research content is both fundamental and timely.

The currently operational unified information platform, the Fanwei Platform, primarily utilizes a low-code + custom development model to facilitate the construction of all information platforms within our institute. The Fanwei Platform boasts powerful core functionalities, with the low-code development model encompassing over 90% of daily application scenarios. Additional special functions can be seamlessly integrated into the platform through custom development, supporting various docking functions such as internal and external interfaces.

The key technical difficulties lie in the following aspects:

1. Implementation of the R&D expense aggregation algorithm. There are eight major categories of R&D expenses, each with different aggregation logic, raw data, and processing logic. It is necessary to conduct in-depth research on the aggregation conditions and raw data processing algorithms under different criteria.

2. A topic identification algorithm suitable for high-tech enterprise recognition and additional deduction. Selecting topics from the overall topic library that meet the conditions for high-tech enterprise recognition and additional deduction requires multi-angle and multi-dimensional logical judgment from the perspectives of enterprise financial data, topic technology information, and financial data, combined with the latest policy logic to make reasonable judgments.

3. Algorithm for improving OCR recognition rate of original data. The general OCR recognition algorithm has a basically consistent recognition rate for various types of written materials. However, financial supporting materials used for scientific research management generally have regular characteristics. By incorporating a recognition algorithm for commonly used financial bills into the general OCR recognition, data processing efficiency can be effectively improved.

Key technological innovation points:

The information management and control system for high-tech enterprises is designed to effectively meet the compliance requirements for project funding management in high-tech enterprises. It integrates the standard logic, process, and form of R&D expense collection required by high-tech enterprise certification, combines R&D expenses and R&D income requirements with enterprise and project information, achieves policy risk investigation, and substantially improves the efficiency of R&D activities. It effectively resolves the correspondence between research management process materials and high-tech enterprise certification and additional deduction declaration materials, ensuring data separation and material separation under multiple perspectives of project materials, and completely solving the stubborn problem of data reorganization.

The system itself relies on the Fanwei low-code platform, achieving a management philosophy of unified technology and standards. At the same time, the construction of the new system does not require the purchase of hardware equipment such as cloud servers and cloud databases, reducing construction costs. The platform meets the requirements of security level protection management and seamlessly integrates with platform module functions, providing a low-risk and low-integration-cost solution with limited overall technological innovation. The ultimate effect is to achieve a resource sharing, data sharing, and an integrated information construction model that integrates graphics and text, as well as digital and intelligent integration.

Research methods:

The finance department (fund management department) is responsible for the overall planning and design of the project, coordinating research and development (R&D) topics and R&D fund management, streamlining workflows, clarifying involved departments and units, and integrating a set of scientific and feasible information management processes based on the R&D fund management system and process forms. At the same time, it collects opinions from internal and external experts for review, analyzes and demonstrates the details of process forms, and discusses from the perspectives of standard system normalization, practicality, and application depth. For the core content of the topic, such as the logic of R&D expense collection, industry experts are invited to deliberate from multiple perspectives and disciplines, and a complete set of logical algorithms is developed to form an algorithm model. The model's validity and reliability are demonstrated by inputting historical data into it. Finally, the algorithm logical model is compiled into archived materials, and R&D personnel are entrusted to embed the logic into the Pan-Micro platform through secondary development mode to achieve functionality. At the same time, the original data structure system is optimized and integrated, ultimately achieving integration with other platform functions, connecting steps, processes, forms, and reports, and integrating data and logic to form a resource sharing platform for data and logic.

Implementation plan:

I、Propose task requirements to the entrusting developer

Through comprehensive planning and analysis by internal and external expert group members, specific modules and functions of the system are determined, core algorithm models are organized, and requirement design documents such as outline design specifications and detailed design specifications are formed. This work plan is scheduled for 20 days. The system construction includes the following main functional modules:

II. Developer develops the system

Design specifications, technical solutions, and other documents according to requirements, assemble a research and development team, and establish research and development plans, specifications, and job responsibilities. The research and development team must strictly adhere to the time schedule milestones to complete the research and development of functional modules and phase testing work.

III. System Trial Operation Evaluation

After the system is developed, it will undergo online deployment and training, followed by a formal trial run period. The trial run period is a crucial time for system operation. This stage serves as both an effective verification period for the design prototype outlined in the requirements specification and a key stage for validating the theoretical model developed in the early stages. All participants involved in the early stages of this stage should fully utilize the system according to the design concept and table specifications, provide feedback, and establish an effective management mechanism. Specifically, the process involves proposing feedback, managing feedback, revising feedback, updating programs, and releasing updated content. Specific tasks require the preparation of a trial run management plan to clarify training plans, communication channels and mechanisms, feedback plans, trial run evaluation plans, trial run results, and other management documents

In recent years, construction enterprises have seen a growing increase in credit balances, leading to prominent tax-related issues in projects. This is manifested in the fact that during the project bidding and budgeting phase, there is only financial budgeting without reasonable tax planning. After commencement, the distribution of input invoice tax points is not reasonable, resulting in either excessive prepaid value-added tax credits or excessive input invoice credits, which has caused certain obstacles to the enterprise's cash flow. To effectively and reasonably manage construction projects, provide reasonable suggestions and guidance from a tax-related perspective, conduct value-added tax planning before project commencement, and perform target comparison and risk warnings after project commencement. Based on the above management requirements, this product draws on the experience of construction enterprises in tax planning from the reform of replacing business tax with value-added tax over the years, conducts in-depth research on the latest policies and regulations, and explores optimization algorithms for the proportion structure of the main costs "labor, materials, and machinery". With the tax burden target of "output tax - input tax = prepaid value-added tax", it integrates information technology to achieve a full-process solution to value-added tax-related issues before and after project commencement.

In recent years, the credit balances of construction enterprises have been increasing, and tax-related issues in projects have become prominent. This is manifested in the fact that during the project bidding and budgeting stage, there is only financial budgeting without reasonable tax planning. After the commencement of construction, the distribution of input invoice tax points is not reasonable, resulting in either excessive prepaid value-added tax credits or excessive input invoice credits, which have caused certain obstacles to the enterprise's cash flow. To effectively and reasonably manage construction projects, provide reasonable suggestions and guidance for enterprises from a tax-related perspective, conduct value-added tax planning before project commencement, and provide target comparison and risk warnings after project commencement. Based on the above management requirements, this product draws on the experience of construction enterprises in tax planning from the reform of replacing business tax with value-added tax over the years, conducts in-depth research on the latest policies and regulations, and explores optimization algorithms for the proportion structure of the main costs "labor, materials, and machinery". With the tax burden target of "output tax - input tax = prepaid value-added tax", it integrates information technology to achieve a full-process solution to value-added tax-related issues before and after project commencement.

Labor subcontracting and professional subcontracting are common in construction enterprises' engineering projects, and prepaid value-added tax can be deducted from subcontract payments.

Before providing a quotation to the owner, with expected revenue and costs already determined, differences in invoicing tax rates for labor, materials, and machinery costs will lead to significant differences in input tax, prepaid value-added tax (VAT), and actual payable tax amounts, having a substantial impact on after-tax profits (VAT and surcharges).

This system automatically calculates several schemes with actual tax payable amount -> 0 by inputting project investment, labor, machinery, materials, and other expenses, providing the financial personnel of the project department with options that align with the actual procurement invoicing situation for selection.

The scheme has three forms, tailored for different user groups to facilitate information viewing.

The company's main business scope includes housing construction engineering, mineral resources development, metallurgical electromechanical equipment installation, municipal public works, etc. Currently, the company holds honorary titles such as Beijing Enterprise Technology Center, Beijing Intellectual Property Demonstration Unit, Science and Technology Innovation Enterprise in Beijing Economic and Technological Development Zone, High-tech Enterprise, Hebei Enterprise Technology Center, and MCC Mining Engineering Center. It has successively undertaken major science and technology projects from the Ministry of Science and Technology, the Ministry of Finance, the Beijing Municipal Science and Technology Commission, and China Minmetals Corporation. The company was recognized as a high-tech enterprise in 2020 and underwent a re-examination and re-recognition in 2023. To strengthen the management of science and technology research and development projects, enhance the company's independent innovation capability, and further maintain its status as a high-tech enterprise, the company developed and launched a science and technology research and development project management platform system.

The technology R&D project management platform has achieved full-process management of scientific research projects, from initiation, implementation to conclusion. Through the project dashboard, the progress of the project can be viewed in real-time, and functions such as managing the expert database and achievement repository are also implemented.

In recent years, the prepaid tax amount of the Shanghai branch has been increasing. Due to the lack of calculation for project input tax certification and prepayment, the input tax certification amount is relatively large, while the prepaid tax amount cannot be offset against the output tax, resulting in a cumulative increase in the balance of prepaid tax. On the other hand, because the input tax certification was not planned in accordance with the content of the Announcement of the State Taxation Administration on Matters Relating to the Handling of VAT Ending Retained Credit Tax Refunds (Announcement No. 20 of 2019), the input tax certification fluctuated, and the tax refund was not processed in a timely manner, leading to increased capital occupation.

To effectively and reasonably manage the withholding tax and tax refund matters of construction projects, provide reasonable suggestions and guidance for enterprises from a tax-related perspective, and specially carry out monthly tax planning for the value-added withholding tax and input tax credit amount of the project, thereby fully utilizing the withholding tax and increasing the tax refund amount, improving the company's cash utilization efficiency and value-added tax management level.

Based on the above actual situation and management requirements, our company has established a project team to conduct in-depth research on the latest policies and regulations, as well as on the optimization algorithm for prepaid taxes and reasonable amounts of value-added tax (VAT) certification. The tax burden target is to reduce prepaid taxes and achieve tax refunds of 20 million to 50 million yuan annually. Integrating information technology, the team focuses on controlling VAT input tax and prepaid tax credit amounts, aiming to achieve a full-process solution for VAT-related issues.

Both prepaid value-added tax and input tax can be offset against output tax, but disorderly offsetting will result in an increasing balance of prepaid and input taxes, leading to a

By collecting historical data and projecting data for future project departments, we can calculate the sales, purchases, and prepaid amounts for each month in the future.

Utilizing the value-added tax deduction algorithm model, we can accurately calculate the deduction relationship between the output tax and the three balances for a certain future month, thereby achieving incremental credit refunds.