Method and system for uploading documents

Abstract

The disclosure relates to a method and system for capturing, uploading and processing documents using an intelligent assistance system. The method comprises the following steps: registering a user on an end user device that is connected to a server using data technology; assigning document attributes; capturing a document, wherein the assistance system checks the document and, in response, transmits a signal or message to the end user device indicating whether the captured document is sufficient and whether a new capture request needs to be provided in the event of an insufficient document; signing the captured document if the latter has been recognized as sufficient by the assistance system; sending the captured signed document to the server; and checking the captured signed document, wherein the document is released for processing if the check is positive.

Description

[0001] DOCUMENT UPLOAD PROCEDURE AND SYSTEM

[0002] The present invention relates to a method for capturing, uploading, and processing documents. Furthermore, the present invention relates to a mobile device for implementing the method and a computer system.

[0003] Technological background

[0004] The areas of document capture, upload and processing are integral components of the broader area of ​​document management systems (DMS).

[0005] Document capture encompasses the process of capturing, digitizing, and converting documents into a format that can be effectively used, managed, and stored. With the advent of OCR (Optical Character Recognition), ICR (Intelligent Character Recognition), and IDR (Intelligent Document Recognition) technologies, this field has gained unprecedented importance. They enable computers to recognize text and characters from a variety of sources, including scanned documents, photos, and even handwriting, representing a significant leap over traditional manual input methods. Not only does the automation of the process save time and resources, but the digital transformation of documents also improves the accuracy and reliability of the data.In addition, digitizing documents makes access to and searching for information much easier, resulting in improved business process efficiency and productivity.

[0006] The upload process, which bridges the gap between capture and processing, involves transferring the captured digital documents to a server or cloud-based storage system. This process has been revolutionized by advances in network technology, cloud computing, and secure encryption methods. It is now possible to upload large volumes of documents quickly and securely, saving companies time and resources while reducing the risks of data breaches.

[0007] At the same time, the use of advanced encryption methods during the upload process enables a higher level of security. This allows companies to minimize the risk of data loss or breaches, which could potentially lead to significant financial losses or legal consequences. Modern encryption technology not only offers the ability to protect documents during transmission but also ensures that the documents are securely stored in the cloud.

[0008] In addition to the encryption technology already mentioned, certificates and digitally signed documents have taken on an important role in the modern upload process to ensure security and authenticity. Digital certificates are a type of "digital ID" used to verify the identity of a user or system during the data transfer process. They are issued by a trusted third party, known as a certificate authority, and can be used to confirm that the data originates from the stated source and has not been tampered with during transmission. In this sense, they help ensure the integrity of the transmitted data and increase trust in the upload process.

[0009] In the final phase, document processing, the uploaded documents are categorized, indexed, and stored for easy retrieval. Computer vision and machine learning are used here to automatically extract relevant data and metadata, perform semantic analysis, and intelligent document classification. Furthermore, natural language processing has introduced features such as voice analysis and automatic summarization, enabling companies to derive meaningful insights from their documents.

[0010] However, as promising as these advances may be, challenges remain. Due to the sensitive nature of some documents, privacy and security concerns remain paramount. Interoperability issues can also arise due to the different document formats and systems. Furthermore, despite the power of modern OCR and NLP techniques, accurately recognizing and processing complex layouts or handwritten text remains a challenge.

[0011] CH714700B1 is known from the prior art. It discloses a computer-implemented method and system for the efficient capture of personal data and associated documents. It involves linking a data object and a personal identifier within a computer system. A visual code is generated that encodes reference information for identifying the linked data object. A visual representation of the code is created. The reference information encoded in the visual code is captured and stored in a mobile communications device using its camera. Subsequently, a data packet is generated in the mobile communications device containing the captured reference information and an image of a document belonging to the person. This data packet is then transmitted to the computer system.After receiving the package, the computer system saves the image of the document and links it to the data object based on the reference information. There is no provision to prevent document modifiability. Data integrity and quality cannot be guaranteed. Furthermore, no systemic support is provided.

[0012] Furthermore, US2013083362 presents an information processing device that is efficient in file processing and storage. The device consists of a first controller that acquires a file and a second controller that receives region information indicating the location of the file on a specific storage medium. Furthermore, the device is equipped with a generator controller that creates a second file from the original and inserts an additional section to accommodate further data. The region information from the second controller is saved as additional data in the new file through the activities of a storage controller. This region information can be stored as encoded data in a thumbnail area as a reduced image or even communicated with external storage devices for file transfer.Interestingly, the second file can be created in the same or a different format than the first file, with formats such as JPEG, PDF, and TIFF supported. Reducing the data volume in the second file and capturing image data to create the first file are also possible. Furthermore, the disclosure includes not only the device itself, but also a non-transferable, computer-readable storage medium containing an information processing program for performing these functions, as well as a method for performing the file processing process.

[0013] US2016253161 relates to an information processing device that interacts with two servers and an external device. The device receives access information and other information from a target object and sends it to the first server. In return, the server provides unique identification and access information for the second server. This information is stored on the device, which then requests access to the second server to download and install an application program. Once the application is installed, the device can receive instructions for its activation. The stored identification information is sent back to the first server to obtain unique information for this identification. This allows the application program to establish a connection with an external device and perform communication processing.The invention further includes a browser program, image processing, wireless near-field communication, a code scheme reader, device control methods, and a communication system. US2017094101 describes a management system that facilitates communication between an image processing device, a terminal, and the system itself. It includes a network interface and a controller. The controller receives position information requests from the image processing device, transmits original image position information, stores the original image data, and transmits code data to acquire target image position information. The system then receives target image data requests from the terminal and transmits the corresponding data. Other functions include generating position information, deleting data, handling various network interfaces, converting image data, and displaying interactive page data.The invention also includes a communication system and a computer-readable storage medium with program instructions.

[0014] Description of the invention

[0015] An object of the invention is to avoid at least some of the disadvantages of the prior art.

[0016] This problem is solved by the features of the independent patent claims.

[0017] The inventive solution comprises a method for capturing, uploading, and processing documents. The method comprises the following steps: logging in a user on an end-user device that is connected to a server, capturing a document, performing document attribution, signing the captured document, sending the captured signed document to the server, and verifying the captured signed document. If the verification is positive, the document is released for processing.

[0018] In a further aspect, a method for capturing, uploading, and processing documents using an intelligent support system is disclosed. The method comprises the following steps: logging in a user to an end-user device that is data-technically connected to a server; performing document attribution; capturing a document, wherein the support system performs a check of the document and, in response, transmits a signal or message to the end-user device indicating whether the captured document is sufficient and, in the case of an insufficient document, whether a new capture request should be provided; signing the captured document if it has been recognized as sufficient by the support system; sending the captured signed document to the server; and verifying the captured signed document, wherein, if the check is positive, the document is released for processing.

[0019] In a preferred embodiment, it is determined whether information about the captured document is sufficient, and if insufficient information is present, an input request is provided. The insufficient information may be based on an insufficient information fragment.

[0020] It is advantageous if an analysis with interpretation is performed for each procedural step, with a support system-dependent interaction taking place. The interaction can include circumstance-driven support. The interaction can include validation of a document to be captured and / or a captured document and / or a captured signed document. For example, missing documents can be detected or inputs validated. A document can be interpreted, and follow-up steps can be planned or initiated based on the document or document content.

[0021] When a document is captured, context-based human-machine support can be used to check the document and, if the document is unsuitable, an error message and / or a new capture request can be issued.

[0022] For example, if an incorrect receipt is uploaded or content is not readable or difficult to read, the support system can send a message or instructions to the end user device, e.g. certain information can be accepted or additional information can be requested.

[0023] The support system can detect poor quality documents or images and ask a user to upload them again.

[0024] Supported or context-based guidance through the process steps, including quality assurance at each step, allows for optimally guided, "systemic" support. This optimal support is further enhanced by technologies such as OCR, AI, LLM, etc.

[0025] Document attribution can be performed by providing one or more selection lists. A client, year, canton, or a category such as a salary statement, annual financial statement, balance sheet, account sheets, contract, etc., can be specified to facilitate appropriate assignment. Document attribution can be performed sequentially or independently of time.

[0026] This method can also incorporate the integration of artificial intelligence. This can help improve the document capture and review process, for example, by using automatic character recognition (OCR) or machine learning for document classification. These technologies could significantly increase the accuracy and speed of document processing. The support system can be implemented as an artificial intelligence (AI)-based system. The support can be provided as associated or context-based support. The intelligent system is self-learning and can utilize neural networks. Through feedback between the AI ​​and a user interface (UI) of the end-user device, complex logic can be simplified and presented visually in a more perceptible way for the user.In short, complexity is reduced and interaction with the class allows for efficient or reduced communication.

[0027] Conventional AI usually only provides answers or feedback and thus there is only one interaction.

[0028] The support system advantageously includes an AI control loop. A user interface (UI) and the AI ​​are in a true feedback loop, meaning that input can be analyzed and controlled by the AI ​​(loop). The AI ​​can suggest improvements to the user on the UI, and the user can make adjustments, which the AI ​​or the support system then checks and, if necessary, further improves or refines in the sense of a control loop.

[0029] A Kl / Ul control loop or simply Kl control loop goes beyond a normal interaction and increases the quality of the achievable results and improves user interaction.

[0030] Feedback on a captured and uploaded document is provided based on the quality of the upload, which is enhanced and not just a transmission confirmation. For example, the document categorization can be reviewed and marked as correct, incorrect, or deficient. For example, if the user entry "salary statement" matches the automatic category recognition, does the document refer to the correct person? Or a completeness check is performed if, for example, page 2 is missing from a three-page document. The process can also be structured in multiple stages.

[0031] For example, a requirement to comply with the terms of the contract can be provided directly with a scannable QR code, with the requirement that the user can capture and upload the document directly and easily at any time or later. For documents requiring evidentiary access, codes should be unique and / or the process should be authorized using a second-factor authentication method.

[0032] Processing using AI-enabled document support systems can encompass various technologies and methods that enable documents to be efficiently scanned, analyzed, and relevant information extracted. Optical character recognition (OCR) enables the scanning and conversion of printed or handwritten text into machine-readable text. Natural language processing (NLP) with NLP algorithms can analyze texts to extract context, meaning, and relevant information. This can be done using Named Entity Recognition (NER), sentiment analysis, and text classification. Machine learning, large language models, or ML models, are used to recognize and extract specific information from documents, such as addresses, invoice numbers, dates, etc. Libraries such as TensorFlow and PyTorch can be used for this purpose.For document analysis, various AI methods can be combined to analyze documents accordingly. Knowledge graphs and semantic analysis can be used to understand and represent relationships between different pieces of information within a document or between different documents.

[0033] The support system includes components of artificial intelligence (Kl), for example to perform a document review.

[0034] By combining technologies and targeted AI support, the efficiency and accuracy of document processing and analysis can be significantly increased.

[0035] The solution according to the invention can be supplemented and further improved by the following additional, each advantageous embodiment.

[0036] According to one embodiment, the method further comprises categorizing the document, the captured document and / or the captured signed document based on the document attribution.

[0037] Incorporating categorization of the document, the captured document, and / or the captured signed document into the process offers several advantages.

[0038] First, this categorization enables improved document organization. It makes it easier for users to find specific documents from a multitude of files and reduces the time required to access these documents.

[0039] Second, categorization increases the efficiency of document processing. By grouping similar documents together based on their attributes, certain processing tasks, such as extracting specific information or performing specific analyses, can be performed more easily and quickly. Third, categorization could help improve data analysis and reporting. With categorized documents, users can better identify trends and patterns within specific categories of documents.

[0040] Additionally, machine learning and artificial intelligence can be integrated into document attribution, for example, to perform document classification using an advanced system. This system facilitates the clear assignment of documents and enables thematic grouping and prioritization.

[0041] For example, a document marked as confidential can be automatically saved to a secure storage area based on its document attribute. Documents related to a specific project could be automatically assigned to the corresponding project folder. This significantly simplifies organization and access to relevant information.

[0042] Furthermore, an algorithm could be developed that analyzes the content of a captured document and automatically assigns a set of relevant document attributes. These could include aspects such as the document subject, the individuals or companies involved, the type of document (e.g., invoice, contract, technical specification), and the time horizon.

[0043] Furthermore, document attributes could be used to indicate the document status, such as whether it is a draft or final version, or whether the document has been approved or still requires approval. This would enable more efficient and transparent document management.

[0044] Additionally, document attribution offers the possibility of implementing a more efficient search function that could search not only for document titles, but for all document attributes. This could significantly increase the efficiency of document searches, especially in large or complex document archives.

[0045] Finally, such a system could also send notifications when documents are created, modified, or approved based on the document attributes assigned to them. This would improve collaboration and communication within teams and departments.

[0046] According to a further embodiment, the capture of a document includes image processing, in particular perspective correction and / or contrast optimization and / or data extraction. This can improve the captured document to enable optimal further processing. Data extraction allows more precise assignment and storage, e.g. in a database. Advantageously, data can be inserted or added to designated fields through automation or verification, e.g. using different colors, which makes it easier to differentiate between the data. Fields can be assigned a data origin, which can be helpful for post-processing. An error assistant can check the extracted data and mark any inconsistencies.

[0047] In addition, user profiles could be created to tailor the document capture and processing process to individual user requirements. For example, personal preferences regarding contrast optimization or perspective correction could be taken into account.

[0048] For data review, in addition to an error assistant, a real-time feedback system could be implemented that immediately detects inconsistencies and alerts the user. The feedback system could also provide suggestions for correcting the inconsistencies to optimize the post-processing process.

[0049] A data validation system could also be introduced to ensure that the extracted data is accurate and complete. Such a system could be based on algorithms that check the data for consistency and plausibility and make automatic corrections if necessary.

[0050] A user can log in using a biometric login, enabling user-friendly and identity-verified login. Biometric data is unique to each individual and includes characteristics such as fingerprints, iris patterns, voice patterns, facial structure, and even behavioral patterns such as typing speed or movement patterns.

[0051] When a user wishes to log into a system, they might be asked to provide a biometric feature, such as a fingerprint or facial recognition. A specialized scanner would then scan or capture this feature and compare the obtained data with a previously stored reference of that feature. If the captured data and the stored reference match, the user is successfully authenticated and granted access to the system.

[0052] To further increase security, multi-factor authentication could be implemented. This method requires the user to provide an additional authentication factor in addition to biometric authentication, such as a password, biometric data, or a code sent via SMS.

[0053] Using biometric data for authentication offers a high level of security, as biometric characteristics are difficult to forge or steal. At the same time, it can be very convenient for users, as they don't have to remember complicated passwords.

[0054] According to a further embodiment, the document attribution is carried out in response to the registration.

[0055] By enabling document attribution immediately after logging in, seamless and rapid document editing is enabled. This can help minimize workflow delays while increasing user experience.

[0056] Furthermore, early document attribution increases the accuracy and relevance of the captured data. It allows for the collection of relevant information about the document and its context, enabling more accurate categorization and processing. The more accurate and relevant the data, the higher the quality of subsequent processes, such as data extraction or image processing.

[0057] Another advantage of this approach is that it allows for process personalization. By performing document attribution based on the logged-in user, the system can tailor the process to the user's individual needs and preferences. This can be used, for example, to group the list of documents by document attributes, such as client and / or tax year, and to offer additional filtering and grouping options.

[0058] According to a further embodiment, the method further comprises querying documents stored on the server.

[0059] This provides a comprehensive picture or overview of already uploaded documents to the end user device and thus helps to avoid duplication.

[0060] Furthermore, by querying documents stored on the server, users can quickly and easily access their documents without requiring additional requests or tools. This saves time and simplifies the process, especially for users who need to access certain documents regularly. This feature also improves system efficiency. Instead of having to re-upload or edit documents, users can simply access existing documents. This means less traffic is generated and the server can operate more efficiently. Furthermore, this feature could also improve system performance by minimizing redundant processes and thus saving resources.

[0061] Furthermore, this feature enables better document management. With the ability to query documents stored on the server, users can more easily organize and manage their documents. For example, they could mark specific documents for later use or filter documents by various criteria, such as creation date, author, or file type.

[0062] According to a further embodiment, it is provided that the document and / or captured document and / or captured signed document are / will be stored temporarily on the end user device.

[0063] “Transient” usually means that the data is stored only temporarily and often only for the duration of a session, similar to the concept of cache memory or temporary files in computer technology.

[0064] Firstly, this feature can ensure that the documents are no longer accessible to unauthorized persons if, for example, the end user device is a mobile device that has been lost.

[0065] First, this method increases system security. By storing documents volatilely, the risk of them being accessed or misused in the event of a device loss or theft is minimized. Once the user ends their session or restarts the device, the volatile data is typically deleted, significantly reducing the risk of permanent data compromise.

[0066] Second, this feature can improve the performance of the end-user device. Because the data is stored only temporarily, less storage space is used on the device. This can help the device run faster and more efficiently, especially on devices with limited storage space.

[0067] According to a further embodiment, after the captured signed document has been sent to the server, the document and / or captured document and / or captured signed document are / will be deleted in whole or in part on the end user device following a positive verification. This ensures that only verified and valid documents are deleted. Partial deletion can be considered for larger documents, thus optimizing memory. This can help optimize device performance and ensure efficient operation, especially when storage space is significantly limited.

[0068] In addition, by deleting the data on the end-user device after successful verification and transfer to the server, the risk of sensitive information falling into the wrong hands is minimized because the sensitive data is no longer present on the device.

[0069] According to a further embodiment, the method further comprises updating a list of documents sent to the server on the end user device.

[0070] With an updated list of sent documents, the user can always keep track of which documents have already been sent to the server. This can help avoid confusion and increase user confidence in the system.

[0071] Furthermore, this feature improves the system's usability. Users have quick and easy access to an overview of all submitted documents, significantly simplifying the search and access process. For example, they can search or filter the list according to various criteria to find specific documents more quickly.

[0072] According to a further embodiment, it is provided that the document and / or captured document and / or captured signed document can contain metadata.

[0073] Metadata can contain a variety of information, such as the document's creation date, location, author, document type, or other specific characteristics. This allows the user or system to quickly obtain relevant information about the document without having to open and read the document itself.

[0074] Second, incorporating metadata can increase the efficiency of the system. Using metadata makes documents easier to categorize, organize, and retrieve. This can greatly facilitate document searches and overall document management.

[0075] According to a further embodiment, after positive verification of the captured signed document, the captured signed document and / or the document attribution and / or the associated metadata are stored on the server. This offers several advantages. First, a separate or external component for signing documents increases the security of the process. Since the signature function is isolated from the main application, this reduces the risk of a comprehensive system compromise. If an attacker gains access to the main system, this separation of components would protect the signature function and the associated sensitive data from direct access.

[0076] Second, this method offers improved scalability. With a separate component for signing documents, the system can be more easily expanded to handle larger workloads without impacting the main application. For example, if the volume of documents to be signed increases, the signing component can be scaled accordingly to handle the increased workload.

[0077] Third, a separate or external component for signing documents brings flexibility to the system. Depending on the specific requirements and circumstances, various external signature services or technologies can be used without significantly changing the main application. This can facilitate adaptation to various legal requirements or technological standards.

[0078] According to a further embodiment, the method further comprises signature storage of the captured signed document on the end user device and / or server.

[0079] First, by storing the signed document, along with its associated metadata and categorizations, on the server, they are protected from local threats on the end-user device. Servers are generally better protected against data loss due to hardware failure and can thus ensure data integrity and availability.

[0080] Second, storing documents along with their categorizations and metadata significantly simplifies data management. Users and system administrators can quickly and efficiently access, filter, and sort the information they need, accelerating workflow and improving the user experience.

[0081] Furthermore, the system's security is significantly increased by storing the signature both locally on the end-user device and on the server, as it allows for the detection of tampering by third parties. In this case, the signature differs between the server and the end-user device, and a document can be declared invalid and thus inconsistent. Furthermore, a copy of the signature is stored on the end-user device to enable traceability of the captured signed documents on the end-user device.

[0082] According to a further embodiment, the verification of the captured signed document comprises a consistency and / or integrity check.

[0083] Data integrity is significantly improved by this verification method. A consistency check can be used to ensure that the document contains no contradictions or inconsistencies that could affect its validity or understandability. An integrity check, on the other hand, ensures that the document has not been altered since its creation or last valid modification. Together, these two checks can help ensure that the data in the captured documents is accurate, consistent, and trustworthy.

[0084] Improved error detection is another benefit of this review method. During the consistency and integrity check, other potential errors or issues in the document can also be identified and highlighted, helping users correct them and improve the quality of the captured documents.

[0085] Ultimately, this type of verification builds confidence in the validity of the documents. By confirming the consistency and integrity of the documents, users can be confident that the information they are viewing or using is accurate and unaltered. This can be critical for a wide range of applications, from managing personal information to complying with legal requirements.

[0086] According to a further embodiment, the verification of the captured signed document takes place on the server.

[0087] First, server-side validation enables better performance and less load on the end-user device. Servers typically have significantly more processing power and resources than end-user devices, enabling them to perform complex validations and calculations more efficiently. This saves time and battery life on the end-user device, enabling a faster and smoother user experience.

[0088] Furthermore, performing the check on the server enables improved scalability. Because the server performs the checks, the system can easily scale to a large number of users and documents without impacting the performance or functionality of the individual end-user device. This is especially important for large organizations or systems with many users.

[0089] Finally, server-side verification also offers enhanced security features. Because the server has control over the verification, it can ensure that the verification processes and results are protected from tampering. Furthermore, the server can also provide better logging and monitoring capabilities to detect suspicious activity and respond to security issues.

[0090] The solution according to the invention also comprises a mobile device that comprises means for carrying out the method.

[0091] Mobility is an obvious and significant advantage of using a mobile device. With the ability to perform the process on a mobile device, users can access and process their documents from anywhere, anytime. This opens up new possibilities for working on the go, at home, or in remote locations, allowing for greater flexibility in the way they work.

[0092] The convenience of using a mobile device is another key advantage. Mobile devices are generally intuitive and offer a user-friendly interface. Furthermore, most people are already familiar with using mobile devices in their daily lives, reducing the learning curve. Furthermore, their size and portability help make the user experience convenient and seamless.

[0093] The solution according to the invention also comprises a computer system that comprises means for carrying out the method.

[0094] Such a computer system increases efficiency because it optimizes work processes and uses resources more efficiently. Focusing on specific tasks allows the system to maximize performance, resulting in faster processing times, fewer errors, and increased productivity.

[0095] Furthermore, the use of a specialized computer system increases the adaptability of the method. It allows the system to be adapted to specific requirements and workflows, thus offering greater flexibility and customized solutions tailored to the individual needs of users. The system can be configured accordingly, depending on whether, for example, the user interface needs to be customized, special reporting functions, or specific processing steps are required. It will be obvious to those skilled in the art that all of the described embodiments can be implemented in an inventive embodiment of the present invention, unless they explicitly exclude each other.

[0096] In the following, the present invention will now be explained in more detail using specific embodiments and figures, without, however, being limited to these.

[0097] By studying these particular embodiments and figures, further advantageous embodiments of the present invention may become apparent to a person skilled in the art.

[0098] Character description

[0099] Embodiments of the invention are described with reference to the following figures.

[0100] Fig. 1 : a diagram of a user login;

[0101] Fig. 2: a diagram of a simplified representation of document capture;

[0102] Fig. 3: a diagram of the requirement to comply with the conditions;

[0103] Fig. 4: a flowchart of the acquisition process;

[0104] Fig. 5: a flowchart for capturing, uploading and processing documents with an intelligent support system;

[0105] Fig. 6 is a schematic representation of a Kl control loop.

[0106] Fig. 7 is a schematic diagram of a system for capturing, uploading and processing documents with an intelligent support system.

[0107] Implementation of the invention

[0108] Figure 1 shows a diagram of a user's login. This represents the first step in the process of capturing, uploading, and processing documents 5. First, the user must be in possession of an end-user device 1, which in this case is a mobile device. The user logs into an application on a server 3. The user is then provided with selection lists that allow them to specify their preferred tax year, a list of mandates, and the desired canton. In this specific example, the user has a choice of several cantons and selects, for example, the canton of Obwalden. Based on the selected canton, an identity provider 4 is called to authenticate the user. The user is then presented with a list of authorized mandates. Once they have selected one, they can use further functions.The entire procedure is adapted to these specific details.

[0109] The identification procedures can vary and are determined by the application operators. In this example, it is two-factor authentication with optional biometric authentication. Finally, after successful identification, a valid certificate 2 is sent back to the user's mobile device 1.

[0110] Figure 2 presents a diagram of simplified document capture. In this process, the authenticated user accesses a cantonal server 3 with their end-user or mobile device 1 to obtain a list of clients. The user can then request and retrieve the associated files for each individual client. Next, the process of scanning a specific document 5 begins—in this example, a salary statement. Using the application installed on the mobile device 1, the salary statement is scanned and then sent to a signing location 6 for signing. Finally, the signed document 5, i.e., the salary statement, is sent back to the cantonal server along with a document attribute 13 containing the document category "salary statement." In this context, the document attribute represents the document category selected by the user.

[0111] Figure 3 illustrates a diagram of the requirement request. The function of the requirement request is to inform the user that not all of the documents required to continue certain processes, such as the assessment of a tax, are yet available. For example, the Cantonal Tax Administration (KSTV) reviews the tax return submitted by the user and determines that certain documents 5, such as a salary statement or Pillar 3a supporting documents, are missing. The KSTV then creates a so-called requirement and requests the missing documents. In an effort to make the process as paperless as possible, all communication should be electronic. To this end, the KSTV informs the user about the creation of the requirement, for example by email or via a push notification directly to the user's mobile phone.

[0112] The user then scans or photographs the requested document 5 and has it signed at a signing location 6. Finally, the document 5 is transmitted to the cantonal server 3, and the KSTV is notified of its submission. Figure 4 shows a flowchart of the capture process, which involves several steps. First, a document 5, for example, a paper-based salary statement, is captured as a digital image, referred to as document capture 7, and temporarily stored. This may involve capturing one or more pages, followed by perspective correction and image optimization with regard to contrast, sharpness, and the like. In addition, the user performs data extraction and, if necessary, a document capture review 8 of the image quality. Once all pages have been captured, the user issues a document capture confirmation 9. The following steps then run automatically.A PDF document is created from the captured image through PDF file generation 10, which is then signed using the digital signing step 11 to create a signed document. This signature can be performed within a signing authority 6. The signed document 5 is then transmitted to the server 3 via a server transmission 12. After successful transmission to the server 3, all data from the capture process (image, PDF, signed PDF) is deleted from the end-user device 1, for example, a tablet. For traceability reasons, the hash of the signature can remain stored on the end-user device 1. Optionally, the document 5 can then be provided with technical information using document attribution 13. In this case, the user selects the document category "salary statement."Document attribution can also be performed outside the scanning application, for example in the associated web application in the browser of another end-user device 1.

[0113] In a further embodiment, the method comprises the following steps: logging in the user, providing selection lists, capturing a document, signing a document, transmitting the document to a server, checking the document by the server, and releasing it for further processing.

[0114] The user logs in to a server-side application. Authentication can be performed via various methods and is usually determined by the application operator. Mobile data, such as location, Wi-Fi information, etc., can be used for security and identification purposes. User identification is relevant for the following steps.

[0115] Selectable lists can be made available to registered users. Examples of selectable lists are clients or tax periods. These are relevant, for example, so that a user can select which client and in which tax year they wish to interact. The above list of selectable lists is not exhaustive; the vast majority of them relate to technical aspects. Capturing a document involves generating a signed PDF document from an image. The image can be captured using a mobile camera directly from the application, or a previously captured image file can be selected. To minimize access to potentially sensitive data on the end user device, all data in a scanning application is stored in a sandbox. All image data is only temporarily stored during the capture process and deleted after capture is complete.A user can capture one or more images (often represented by individual pages of a paper document).

[0116] After selecting or capturing an image, the application optimizes the image to maximize the readability of the content. This can include perspective corrections, rotating and cropping the image content, adjusting the contrast, sharpening the image content, and other optimizations. Once the image content is optimized, it is displayed to the user for review and may require confirmation. Upon successful confirmation, a PDF file is generated from the image, which is then signed.

[0117] Signing a document involves signing the PDF generated in the previous step. The signing can be performed within the application or by an external signatory. The result of the signing is a signed file, for example, a PDF-A document. Thanks to the signature, any changes to the content can be detected. This is relevant in the process to detect any changes by a "man in the middle" on the server side.

[0118] The next step is to transfer the document to the server. The user's initial login to the application allows the scanning application to communicate with the server backend. The signed file is securely transmitted to the server backend via an encrypted channel.

[0119] Finally, the document is reviewed and released for further processing. The server backend checks the received content and returns result-based feedback to the scanning application.

[0120] OK (OK): If the received content is acceptable, the document is released for further processing. The scanning application is notified of successful receipt, after which it deletes the temporarily stored data. A hash of the signed file may be stored on the end user's device for traceability (sandbox). NOK (NOK): If the content is not acceptable, e.g., an invalid signature, the content is discarded on the server side. The scanning application receives appropriate feedback, reacts differently depending on the error, and informs the user.

[0121] Another example of this is capturing a paper document as an electronic attachment to a tax return when completing the tax return. The following explains one example of this use case. User X is a natural person and is completing their tax return (NP). When entering income, the pay slip is specified as a mandatory attachment. However, User X only has the pay slip in paper form. The user is informed that attachments can also be captured using a mobile phone and launches a scanning application.

[0122] I. User X starts the scanning application.

[0123] II. After logging in, the clients for which user X is authorized are displayed.

[0124] III. When the client is selected, the associated tax files are retrieved from the server and displayed in the scan application.

[0125] IV. After selecting the desired dossier, the user can enter attachments.

[0126] V. The user starts the capture process, during which the following actions can occur. a. Capturing the paper pay slip as a digital image (volatile storage) i. Capturing one or more pages ii. Perspective correction Image optimization (contrast, sharpness, etc.) iii. Data extraction b. Any image quality check by the user c. Once all pages have been captured, this is confirmed by the user. The subsequent steps are largely automated. d. A PDF document is then generated from the image. e. The generated PDF is digitally signed, creating a "signed document". i. The signature can be applied within the application or by a signing authority. f. The signed document is transmitted to the server. i. Upon successful transmission to the server, all data from the capture process (image, PDF, signed PDF) is deleted from the mobile device.For traceability reasons, the signature hash may still be stored on the device. g. (optional) The document can then be tagged with technical information. In this case, the user selects the document category "Payroll Statement." Document tagging can also be performed outside of the scanning application, for example, in the associated web application in the browser of another device.

[0127] A secure, user-friendly, and supportive document upload system includes several components to ensure the confidentiality, integrity, and availability of documents.

[0128] Some components are listed below.

[0129] The first component is user authentication and authorization. Authentication ensures that only authenticated users can access the document upload system. This can be achieved through robust authentication mechanisms such as two-factor authentication (MFA). Authorization can be achieved through role-based access control (RBAC) to restrict access based on user roles. Only authorized users are allowed to upload, view, or manage documents.

[0130] Secure communication channels such as SSL / TLS can be used to encrypt data transmitted between an end-user device 1 and a server 5, thus preventing eavesdropping and man-in-the-middle attacks.

[0131] Data validation and sanitization should be performed. File type restrictions can prevent malicious files from being uploaded. For example, only certain file types such as PDFs, DOCX, JPEG, etc. are allowed, and size limits are implemented to prevent denial-of-service attacks. File names and contents should be checked to remove potentially malicious code or scripts.

[0132] Storage security should include encryption. Files should be protected with a strong encryption algorithm. This ensures that even if the storage is compromised, the files remain unreadable without the decryption keys.

[0133] Strict access control ensures that only authorized personnel can access the storage system.

[0134] Real-time scanning of files for viruses and malware during upload allows for reliable detection and removal of viruses and malware. An audit log is maintained for all upload activities, including who uploaded a file, when it was uploaded, and all accesses or changes to the file. Continuous monitoring allows for the detection and response of suspicious activity in real time.

[0135] The user interface and experience are enhanced by a support system (14). Progress indicators, circumstantial support, and support system-driven interactions via dialog provide users with feedback and assistance during or after an upload process. Context-based notifications enable users to proceed efficiently and correctly during document uploads.

[0136] Regulatory compliance ensures that the system complies with relevant regulations and standards, such as GDPR, HIPAA, etc., depending on the type of documents.

[0137] Modern implementation technologies are used. Secure server frameworks and languages ​​such as Node.js, Python (Django), or Java (Spring Boot) should be used as the backend. Databases such as PostgreSQL, MySQL, or MongoDB can be provided with encryption and appropriate access controls. Cloud services or cloud storage solutions such as AWS S3 or Azure Blob Storage are equipped with integrated security features.

[0138] Various tools and technologies can be used for the support system and AI-based processing, such as Tesseract OCR; NLTK and spaCy for natural language processing in Python; OpenCV for image processing; PyTorch and TensorFlow for training and applying machine learning models; and cloud services such as AWS Textract, Google Cloud Vision, and Microsoft Azure Cognitive Services for comprehensive document analysis capabilities.

[0139] In a workflow example, a user logs in. The user authenticates using MFA on their end-user device 1, which is connected to a server 3. This is followed by document attribution 13, before capturing a document 5 and uploading the file. The user selects a file 5 to upload. The support system 14 checks the document 5 and, in response, sends a signal to the end-user device 1 indicating whether the captured document 5 is sufficient. If the document 5 is insufficient, a capture request is issued. The support system 14 checks the file type and size, and sanitizes the input. For security reasons, the file 5 is subjected to a virus scan. The file 5 is scanned for malware. The file 5 is then encrypted and securely stored. At the same time, logging takes place, where an entry is created in an audit log.The user receives confirmation or further assistance with the upload. Implementation provides a secure document upload system that protects sensitive information and enjoys user trust.

[0140] Figure 5 shows a flowchart for capturing, uploading and processing documents 5 with an intelligent support system 14.

[0141] In a first step S1, a user logs in to an end user device 1 that is data-linked to a server 3. In a second step S2, document attribution 13 is performed. In step S3, a document 5 is captured, with the support system 14 checking the document 5 and, in response, transmitting a signal to the end user device 1 as to whether the captured document 5 is sufficient and, if the document 5 is insufficient, whether a capture request should be executed. If the captured document 5 is recognized as sufficient by the support system 14, the captured document 5 is signed in step S4. Subsequently, in step S5, the captured signed document 5 is sent to the server 3. In step S6, the captured signed document 5 is checked, and if the check is positive, the document 5 is released for processing or further processing.

[0142] Fig. 6 shows a schematic representation of an AI control loop. The support system 14 connects the user interface (UI) of the end-user device and the AI ​​to a feedback loop, meaning that an input is analyzed and controlled by the AI ​​(a "loop"). The AI ​​can suggest something to a user on the UI, and the user can make adjustments, which the AI ​​then checks and, if necessary, further improves or refines.

[0143] Fig. 7 shows a schematic representation of a system for capturing, uploading, and processing documents with an intelligent support system 14. An end-user device 1 is connected to a network 15. A server 3, an identity provider 4, and a signing authority 6 are also connected to the network 15. The support system 14 is also connected to the network 15 and can be implemented as a cloud service and / or combined with the server 3. A transmission or server transmission 12 takes place via the network 15. Using the end-user device 1, a document 5 is captured as a digital image, which is referred to as document capture. The support system 14 is K1-based and influences the end-user device 1 and its graphical interface (UI) or help or menu navigation.

[0144] The disclosed methods and system can ensure secure and tamper-proof transmission of documents. The present disclosure relates to a method for capturing, uploading, and processing documents. It goes without saying that numerous other embodiments are conceivable for a person skilled in the art based on the exemplary embodiments described.

[0145] List of reference symbols

[0146] 1 end-user device

[0147] 2 Certificate

[0148] 3 servers

[0149] 4 identity providers

[0150] 5 Document

[0151] 6 Signatory body

[0152] 7 Document recording

[0153] 8 Document Intake Review

[0154] 9 Document Acceptance Confirmation

[0155] 10 PDF file generation

[0156] 11 Digital signing

[0157] 12 Server transmission

[0158] 13 Document attribution

[0159] 14 Support system

[0160] 15 Network

Claims

Patent claims 1. A method for capturing, uploading and processing documents (5) with an intelligent support system (14), the method comprising the following steps: - logging in a user on an end user device (1) that is data-technically connected to a server (3); Carrying out document attribution (13); Capturing a document (5), wherein the support system (14) performs a check of the document (5) and in response transmits a signal to the end user device (1) as to whether the captured document (5) is sufficient and, in the case of an insufficient document (5), a capture request is to be provided; Signing the captured document (5) if it has been recognized as sufficient by the support system (14); Sending the captured signed document (5) to the server (3); and Checking the captured signed document (5), whereby if the check is positive the document (5) is released for processing.

2. Method according to claim 1, further comprising: performing an analysis with interpretation for the respective method step, wherein a support system-related interaction takes place.

3. The method of claim 2, wherein the interaction comprises circumstance-controlled support.

4. The method according to claim 2, wherein the interaction comprises a validation of a document (5) to be captured, and / or a captured document (5), and / or a captured signed document (5).

5. The method according to claim 1, further comprising: based on the document attribution, a categorization of the document (5), and / or captured document (5), and / or captured signed document (5).

6. Method according to one of claims 1 to 5, wherein the capturing of the document (5) comprises image processing, in particular perspective correction and / or contrast optimization and / or data extraction.

7. Method according to one of claims 1 to 6, wherein the document attribution (13) is carried out in response to the registration.

8. The method according to any one of claims 1 to 7, further comprising: querying documents (5) stored on the server (3).

9. Method according to one of claims 1 to 8, wherein the document (5) and / or captured document (5) and / or captured signed document (5) are / will be stored temporarily on the end user device (1).

10. The method according to one of claims 1 to 9, wherein after sending the captured signed document (5) to the server (3), the document (5) and / or captured document (5) and / or captured signed document (5) are / will be deleted in whole or in part on the end user device (1) after a positive check.

11. The method according to any one of claims 1 to 10, further comprising: updating a list of the documents (5) sent to the server (3) on the end user device (1).

12. Method according to one of claims 1 to 11, wherein the signing of the captured document (5) comprises metadata.

13. Method according to one of claims 1 to 12, wherein the signing of the captured document (5) is carried out by a separate component or an external component.

14. Method according to one of claims 1 to 13, wherein after positive verification of the captured signed document (5), the captured signed document (5), and / or the categorization, and / or the associated metadata are / is stored on the server (3).

15. The method according to any one of claims 1 to 14, further comprising: signature storage of the captured signed document (5) on the end user device (1) and / or server (3).

16. Method according to one of claims 1 to 15, wherein the verification of the captured signed document (5) comprises a consistency and / or integrity check.

17. Method according to one of claims 1 to 16, wherein the verification of the captured signed document (5) takes place on the server.

18. Method according to one of claims 1 to 17, wherein the support system (14) comprises components of artificial intelligence (Kl).

19. The method according to claim 18, wherein the support system (14) comprises a KL control loop.

20. A mobile device comprising means for carrying out the method according to any one of claims 1 to 19.

21. A computer system comprising means for carrying out the method according to any one of claims 1 to 19.

Images