Bidding book and bidding book automatic generation method based on AI all-in-one machine

Abstract

The invention relates to the technical field of automatic file generation, and provides an AI all-in-one machine-based bid invitation book and bidding book automatic generation method, which comprises the following steps of: 1, filling a company information form by a user; 2, for company information filled by a user, word segmentation processing is carried out on characters, and the characters are stored in a vector database; step 3, in the management page of the bid invitation book or the bidding book, after the bid invitation company uploads project requirements or the bidding company uploads the bidding book, the user sends the project requirements or the bidding book to the local large language model for analysis, and identifies the field to which the project requirements or the bidding book belongs; 4, calling a template of a corresponding field in the bid invitation book template library or the bid invitation book template library, and filling the company information in the vector database into the bid invitation book template or the bid invitation book template of the corresponding field; and 5, obtaining the content filled in the input box by the user, and forming the content of the bid invitation book or the bid tendering book. The writing efficiency of the bid inviting book and the bidding book can be improved, and time is saved.

Description

Technical Field

[0001] The present invention relates to the technical field of automatic document generation, and particularly to a method for automatically generating tender documents and bid documents based on an AI all-in-one machine. Background Art

[0002] The existing AI all-in-one machines on the market are servers with an online large language model connected to the network. If used while traveling, various external devices such as monitors, keyboards, and mice need to be equipped, which is not only very troublesome but also extremely fragile. Once dropped or knocked, they cannot be used, and the cost is very high. At the same time, if the user's data information cannot be connected to the network and data security needs to be protected, or the working environment is relatively remote without signal, the online large language model cannot be used, which does not meet the requirements for other industries with data confidentiality such as the military and scientific research personnel. And to carry a high-quality offline large language model, a very large graphics card is required. Although a desktop computer can install multiple large graphics cards, it cannot be made portable. The maximum graphics card of an ordinary notebook is 16G, so there is no AI all-in-one machine on the market with large video memory + portability + triple solid state + anti-drop.

[0003] For most enterprises, the writing of tender documents and bid documents is a huge project. One person can probably only write one in a week. And if the bid document is modified, the employee needs to rewrite the bid document little by little according to the modified part, which is very troublesome.

[0004] There is no product on the market that can completely generate a tender document from scratch according to project requirements or generate a bid document from scratch based on a tender document. Basically, users have already written a general outline and then assisted in editing, resulting in low writing efficiency of tender documents and bid documents and not solving the real pain points. Summary of the Invention

[0005] The present invention mainly solves the technical problem that there is currently no product that can generate a tender document from scratch according to project requirements or generate a bid document from scratch based on a tender document, resulting in low writing efficiency of tender documents and bid documents. A method for automatically generating tender documents and bid documents based on an AI all-in-one machine is proposed to improve the writing efficiency of tender documents and bid documents and save time.

[0006] The present invention provides a method for automatically generating tender documents and bid documents based on an AI all-in-one machine, including the following processes:

[0007] Step 1, the user fills in the company information form;

[0008] Step 2, perform word segmentation on the text of the company information filled in by the user and store it in the vector database;

[0009] Step 3, after the user uploads project requirements in the management page of the tender document or the bidder uploads the tender document, the project requirements or the tender document are sent to the local large language model for analysis to identify the field to which the project requirements or the tender document belong;

[0010] Step 4, call the templates corresponding to the field in the tender document template library or the bid template library, fill in the company information in the vector database into the tender document template or the bid template corresponding to the field, and display the input boxes that need to be filled in by the user;

[0011] Step 5, obtain the content filled in by the user in the input box to form the content of the tender document or the bid.

[0012] Further, after step 5, it further includes:

[0013] Step 6, embed a word editing plugin in the formed tender document or bid so that the user can modify the format and modify the content of the tender document or the bid.

[0014] Further, the company information includes but is not limited to company basic information, personnel information, winning bid and transaction cases, and similar project cases, and financial data.

[0015] Further, in step 2, for the text of the company information, use natural language word segmentation technology to perform word segmentation processing, split the continuous Chinese character sequence into individual words, and then perform segmentation according to the token length of the sentence to split the company information into text segments.

[0016] Further, the establishment process of the bid template library:

[0017] According to the content format characteristics of the bid content in different fields, extract the common content in each field to form bid templates for multiple fields and store them in the bid template library.

[0018] Further, the establishment process of the tender document template library:

[0019] Parse a large number of tender documents in each field into HTML format; the parsed tender documents are processed by the local large language model to separate the parts that need to be filled in; the parts that need to be filled in include but are not limited to key information such as project requirements, technical solutions, and commercial terms; then these HTML formats are sent to the local large language model again to form tender document templates for each field and store them in the tender document template library.

[0020] Further, the AI all-in-one machine is in the form of a laptop computer; the shell of the AI all-in-one machine uses 3D printing technology; the AI all-in-one machine is equipped with biometric technology.

[0021] A method for automatically generating tender documents and bid documents based on an AI all-in-one machine provided by the present invention has the following advantages compared with the prior art:

[0022] 1. The AI all-in-one machine adopts a rugged three-proof shell design, with features such as anti-drop, waterproof, shockproof, and dustproof. It can be used out of the box without additional external devices, has a large video memory, greatly improves the portability and reliability of the device, and is suitable for application requirements in various harsh environments.

[0023] 2. The AI all-in-one machine is equipped with an offline local large language model and an independent database, ensuring that user data is stored and processed locally throughout the process, eliminating the risk of online data leakage, and is especially suitable for high-security industries.

[0024] 3. Traditionally, it takes about one person-week for an enterprise to write a complete tender document or bid document. If the tender document is modified, the corresponding bid document needs to be adjusted according to the updated points, which is very troublesome. The present invention can generate a tender document or bid document within 30 minutes only by uploading the corresponding materials, greatly saving 90% of the time and improving the writing efficiency of tender documents and bid documents.

[0025] 4. Users can download the tender document or bid document to the local in the word editing plugin. The present invention also supports users to customize templates. Users can upload their own template files, and the backend will also parse the template files into HTML language and render them on the page. Users can edit and fill in according to their own needs and habits. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 is a schematic diagram of the establishment process of the bid document template library;

[0027] Figure 2 is a schematic diagram of the establishment process of the tender document template library;

[0028] Figure 3 is a flowchart of the implementation of the method for automatically generating tender documents and bid documents based on an AI all-in-one machine provided by the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0029] To make the technical problems solved, the technical solutions adopted, and the technical effects achieved by the present invention clearer, the present invention will be further described in detail below with reference to the drawings and embodiments. It can be understood that the specific embodiments described herein are only used to explain the present invention, rather than limiting the present invention. In addition, it should be noted that for the convenience of description, only part of the content related to the present invention is shown in the drawings rather than all of it.

[0030] A method for automatically generating tender documents and bid documents based on an AI all-in-one machine provided by an embodiment of the present invention.

[0031] The AI all-in-one computer is a ready-made laptop that can be used out of the box without any external devices. The shell of the AI all-in-one computer is made by 3D printing technology to create a hard shell with a shape that is resistant to falling, waterproof, shockproof and dustproof, making the product not easily damaged in various harsh environments and can be used anytime and anywhere. The AI all-in-one computer is equipped with biometric recognition technology, providing better privacy protection; it has a super large 4090 24G turbo graphics card + 64G super large memory and an i9 processor. The AI all-in-one computer is equipped with an offline local large language model, supporting the intelligent writing of bidding documents, giving full play to the advantages of AI. Users can upload project requirements / tender documents, and a high-quality tender document / bidding document that meets the requirements can be generated from scratch in a short time, and the file can be downloaded. All the generated data will be saved locally and not shared with the outside world. It truly is an AI all-in-one computer that supports the writing of tender documents and bidding documents, with a large video memory, portability, triple protection and anti-interference, and has extremely high cost performance.

[0032] The present invention requires the use of a bidding document template library and a tender document template library. The tender document template library stores tender document templates and bidding document templates in multiple fields. Fields such as computer, architecture, finance, and other industries each have their own templates in their respective fields.

[0033] As Figure 1 shown, the establishment process of the tender document template library is as follows:

[0034] According to the content format characteristics of tender documents in different fields, through a large amount of accumulation and summary, the common content in each field is refined to form tender document templates in multiple fields, which are stored in the tender document template library. The fields include computer, architecture, finance, and other industries.

[0035] These tender document templates cover the key elements and structures of tender documents in different fields. As the basic framework of tender documents, the tender document templates are then parsed into HTML language for rendering on the page to display the format and structure of tender documents, and are stored in the tender document template library. Users can create custom tender document templates and store them in the tender document template library.

[0036] As Figure 2 shown, the establishment process of the bidding document template library is as follows:

[0037] First, a large number of tender documents in various fields are parsed into HTML format; the parsed tender documents are processed by a local large language model to separate the parts that need to be filled; the parts that need to be filled include but are not limited to key information such as project requirements, technical solutions, and commercial terms; then these HTML formats are sent to the local large language model again to form a tender document template for each field and stored in the tender document template library. The tender document template for each field covers the common requirements and format requirements within the field. Users can create custom tender document templates and store them in the tender document template library.

[0038] As Figure 3 shown, a method for automatically generating tender documents and bid documents based on an AI all-in-one machine provided by an embodiment of the present invention includes the following processes:

[0039] Step 1, the user fills in the company information form.

[0040] The user is a tendering company or a bidding company. The company information includes but is not limited to basic company information, personnel information, winning bid and transaction cases, and similar project cases, financial data, etc. It is used as an information introduction for the tendering company or as basic filling content for the bidding company.

[0041] Step 2, for the company information filled in by the user, the text is segmented and stored in the vector database.

[0042] For the text of the company information, natural language segmentation technology is used for segmentation. The continuous Chinese character sequence is segmented into individual words, and then segmented according to the token length of the sentence to segment the company information into text fragments.

[0043] The result after segmentation captures the semantic relationship between words. Through the vocabulary table and vector mapping, the text is converted into a semantic relationship vector and stored in the vector database.

[0044] Step 3, in the management page of the tender document or bid document, the user clicks "New". After the tendering company uploads the project requirements or the bidding company uploads the tender document, the project requirements or the tender document are sent to the local large language model for analysis to identify the field to which the project requirements or the tender document belong.

[0045] The tendering company uploads the project requirements to automatically generate tender documents. The bidding company uploads the tender document to automatically generate bid documents.

[0046] Large Language Model (LLM) refers to a deep learning model trained using a large amount of text data, which can generate natural language text or understand the meaning of language text. Large language models can handle a variety of natural language tasks, such as text classification, question answering, dialogue, etc., and are an important way to artificial intelligence.

[0047] Step 4: Call the template in the corresponding field in the tender document template library or the bid document template library, fill in the company information in the vector database into the tender document template or the bid document in the corresponding field, and display the input box for the user to fill in.

[0048] For example, automatically fill in company introductions, financial data, etc. in the tender document template or the bid document template in the corresponding field. For example, generate the initial content of the tender document, and the generated content will be rendered in HTML format on the left side of the page. The user can perform AI-assisted filling on the right side. The AI will provide corresponding filling suggestions and reference content according to the information input by the user and the template requirements.

[0049] For example, generate the initial content of the bid document, and the generated template will be HTML-rendered on the left side of the page. The content to be filled will be highlighted. The user can fill in the specific content in the lower right corner of the page, and the reference opinions given by the local large language model based on the general knowledge base will be in the upper right corner.

[0050] Step 5: Obtain the content filled in by the user in the input box to form the tender document or the bid document content.

[0051] Step 6: Embed a word editing plugin in the formed tender document or bid document so that the user can modify the format and the content of the tender document or the bid document.

[0052] The user can download the tender document or the bid document to the local within the word editing plugin. The present invention also supports the user to customize the template. The user can upload their own template file, and the backend will also parse the template file into HTML language and render it on the page. The user can edit and fill it according to their own needs and habits.

[0053] The user can view the list of generated tender documents or bid documents. Click New, upload the tender document and the user's company information, and a bid document template can be generated within twenty minutes. The progress can be viewed by returning to the management page during the generation process. After the bid document is generated, not only can the content be filled in the blank spaces, but also the local large language model gives reference opinions, which can be directly filled back if adopted. After all the content is filled in, enter the embedded word editor to adjust the document format, and finally download it to the local.

[0054] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: modifying the technical solutions recorded in the foregoing embodiments, or equivalently replacing some or all of the technical features therein, does not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for automatically generating tender documents and bids based on an AI all-in-one machine, characterized in that: The process includes: Step 1, the user fills in the company information form; Step 2: Segment the text of the company information filled in by the user and store it in a vector database; Step 3: On the management page of the tender document or bidding document, after the tendering company uploads the project requirements or the bidding company uploads the tender document, the project requirements or tender document are sent to the local large language model for analysis to identify the field to which the project requirements or tender document belongs; Step 4, calling the template of the corresponding field in the tender document template library or the bid document template library, filling the company information in the vector database into the tender document template or the bid document template of the corresponding field, and displaying the input box that needs to be filled in by the user; Step 5, obtain the content filled in by the user in the input box to form the content of the tender document or bid document.

2. The method for automatically generating tender documents and bidding documents based on an AI all-in-one machine according to claim 1 is characterized in that: After step 5, it also includes: Step 6, when forming the tender document or bid, embed a word editing plug-in to allow the user to modify the format and content of the tender document or bid.

3. The method for automatically generating tender documents and bids based on an AI all-in-one machine according to claim 1 is characterized in that: The company information includes but is not limited to basic company information, personnel information, successful bid cases and similar project cases, and financial data.

4. The method for automatically generating tender documents and bids based on an AI all-in-one machine according to claim 2 is characterized in that: In step 2, the text of the company information is segmented using natural language segmentation technology to divide the continuous Chinese character sequence into individual words, and then segmented according to the token length of the sentence to divide the company information into text segments.

5. The method for automatically generating tender documents and bids based on an AI all-in-one machine according to claim 1 is characterized in that: The process of establishing the bidding document template library is as follows: Based on the content format characteristics of bidding documents in different fields, common contents in each field are extracted to form bidding document templates for multiple fields, and stored in the bidding document template library.

6. The method for automatically generating tender documents and bids based on an AI all-in-one machine according to claim 1 is characterized in that: The process of establishing the bidding document template library: A large number of tender documents in various fields are parsed into HTML format; the parsed tender documents are processed through the local large language model to separate the parts that need to be filled in; the parts that need to be filled in include but are not limited to key information such as project requirements, technical solutions, and business terms; these HTML formats are sent to the local large language model again to form a tender document template for each field, and stored in the tender document template library.

7. The method for automatically generating tender documents and bids based on an AI all-in-one machine according to claim 1 is characterized in that: The AI ​​all-in-one machine is in the form of a laptop computer; the shell of the AI ​​all-in-one machine adopts 3D printing technology; the AI ​​all-in-one machine is equipped with biometric recognition technology.

Images