Abstract: A machine learning based (ML-based) method and system for automatically categorizing documents, is disclosed. Initially, the documents are obtained from data sources and pre-processed to generate the pre-processed data associated with documents. The documents are classified as at least one of: relevant financial statements and non-relevant financial statements, based on the pre-processed data using a voting classifier with machine learning (ML) models. The classified non-relevant financial statements are classified into the relevant financial statements using at least one of: a rule-based classification technique and a classifier model, to mitigate false positive categorization of the documents. The re-categorized electronic documents are provided as an output, to users on user interfaces associated with electronic devices associated with the users. The financial statements are classified using TF-IDF vectorizer with voting classifier based on contents of the documents.

Abstract: A machine learning based (ML-based) method and system for automatically extracting and correcting financial information from documents, is disclosed. Initially, the documents are obtained from data sources and pre-processed to generate the pre-processed data associated with contents within the document. The contents are classified as potential key-value pairs corresponding to the financial information based on the system prompts and extracted using the ML model. The potential key-value pairs are corrected to obtain the corrected key-value pairs based on custom prompts, using the ML model. The corrected key-value pairs corresponding to the financial information are provided as the output to the end users on user interfaces associated with an electronic device. This technique extracts financial information regardless of structure or alignment by learning to recognize any added or removed prefixes or suffixes, enabling the prefixes or suffixes to make corrections and generate accurate key-value pairs.

Abstract: A system and method for data migration in cloud databases is disclosed, wherein structured and unstructured data are transferred between heterogeneous cloud environments in a controlled, reliable, and automated manner. The system comprises a processing unit coupled with memory and a set of data handling components configured to extract source data from a first cloud database, perform schema interpretation and transformation, and load the processed data into a target cloud database while preserving data integrity, consistency, and transactional relationships. The migration process includes mechanisms for data validation, conflict resolution, incremental synchronization, and fault tolerance to ensure minimal downtime and reduced operational disruption. The method further enables real-time monitoring of migration status, adaptive resource allocation, and automated rollback in the event of failure conditions.

Abstract: A machine learning based (ML-based) method and system for automatically extracting and correcting financial information from documents, is disclosed. Initially, the documents are obtained from data sources and pre-processed to generate the pre-processed data associated with contents within the document. The contents are classified as potential key-value pairs corresponding to the financial information based on the system prompts and extracted using the ML model. The potential key-value pairs are corrected to obtain the corrected key-value pairs based on custom prompts, using the ML model. The corrected key-value pairs corresponding to the financial information are provided as the output to the end users on user interfaces associated with an electronic device. This technique extracts financial information regardless of structure or alignment by learning to recognize any added or removed prefixes or suffixes, enabling the prefixes or suffixes to make corrections and generate accurate key-value pairs.

Abstract: A machine learning based computing method for extracting data from electronic documents, is disclosed. The machine learning based computing method includes steps of: receiving the electronic documents from data sources; converting formats of the electronic documents into first images associated with the electronic documents; classifying the first images associated with the electronic documents as second images including tabular data, and third images including non-tabular data, based on a CNN based machine learning model; determining tabular regions in tabular data associated with the second images by identifying bounding box coordinates indicating tabular regions in tabular data, based on a YOLO architecture based machine learning model; extracting first data from the tabular regions; extracting second data from non-tabular regions; and providing an output of the first data and the second data to second users on a user interface associated with electronic devices.

Abstract: A machine learning based computing method for generating electronic mails for accounts receivables management is disclosed.

Abstract: A machine learning based (ML-based) computing method and system for automatically categorizing electronic documents in electronic mails, is disclosed. Initially, data associated with the electronic documents are obtained from databases. The data are pre-processed to generate pre-processed texts. The pre-processed texts associated with the electronic documents are analyzed to classify the pre-processed texts into one of a finance related content and a non-finance related content, using a ML model. The electronic documents are categorized as one of electronic financial documents when the pre-processed texts are classified as finance related content, and electronic non-financial documents when the pre-processed texts are classified as non-finance related content, using the ML model.

Abstract: Embodiments include herein are directed towards a circuit having a core bandgap reference circuit including a first transistor, a second transistor, and a differential amplifier. The circuit may further include a cascoded bandgap reference circuit in electrical communication with the core bandgap reference circuit, wherein the cascoded bandgap reference circuit includes a first supply noise path that includes a plurality of transistors that are electrically connected and a flipped voltage follower configuration that reduces an impedance at a transistor associated with a second supply noise path.

Abstract: The present disclosure provides a system and a method for performing connectivity data flow in a network is described. A backhaul router (302) is configured for plurality of virtual local area networks (VLANs) to support indoor small cell (IDSC) connectivity for at least one of mobile network generation by creating a first set of VLANs is created in the backhaul router (302) and a 5G IDSC (300) associated with a first one of mobile network generation signalling/data packets, a second set of VLANs is created in the backhaul router (302) and a 4G IDSC (414) associated with a second one of mobile network generation signalling/data packets and bridging the second set of VLANs is bridged to the 5G IDSC (300) to allow mobile network generation packets of second one of mobile network generation from the backhaul router (302) to the 4G IDSC (414).

Abstract: The present disclosure provides a system and a method for network optimization. In particular, the present disclosure provides a fifth generation (5G) indoor small cell (IDSC) (604) with daisy chain to support backhaul connectivity of 4G combo IDSC (606), consisting of 4G radio access network (RAN) and Wireless-Fidelity (Wi-Fi) access point. The 5G IDSC (604) has one dedicated Ethernet/Optical port to connect the 4G combo IDSC (606). The 5G IDSC (604) bridges the data, control, and precision time protocol (PTP) signals from backhaul router (602), connected to 5G IDSC (604) Ethernet/Optical port, and re-route it to 4G combo IDSC (606) through dedicated Ethernet/Optical port.

Abstract: Embodiments included herein are directed towards a ring voltage-controlled oscillator based phase locked loop circuit. Embodiments may include an integral path including a plurality of resistors and a plurality of transistors configured to correct low frequency variations and a proportional path in electrical communication with the integral path, the proportional path including a plurality of transistors and a source degeneration circuit. Embodiments may further include a decoupling capacitor located at a ring voltage-controlled oscillator node in electrical communication with the proportional path, wherein the source degeneration circuit operates to cancel any effect from the ring voltage-controlled oscillator node.

Abstract: A fiber reinforced composite material comprising the following components: (i) polyacrylonitrile (PAN) particles containing an acrylonitrile content of 80-100 mol % and present in an amount of 0.001-15 wt % by weight of components (i) and (ii); and (ii) a nitrile-containing polymer having a composition different than the PAN particles; wherein nitrile groups in component (i) are crosslinked with nitrile groups of component (ii), and wherein said PAN particles are dispersed in component (ii). Methods for producing the composite material, molded forms thereof, and articles thereof, are also described.

Abstract: A machine learning based computing method for analyzing intent of electronic mails, is disclosed. The machine learning based computing method includes steps of: receiving metadata associated with the electronic mails from databases; extracting first textual contents from last received electronic mails stored in electronic mail files by preprocessing the last received electronic mails; analyzing the intent of the electronic mails based on a first machine learning model; classifying the electronic mails into reason codes based on a second machine learning model; extracting information associated with named entities from unstructured and unlabeled electronic mails based on a third machine learning model; grouping the electronic mails into categories based on the intent of the electronic mails, the reason codes, and standardized information associated with the named entities; and providing an output of categorized electronic mails to users on a user interface associated with electronic devices.

Abstract: A machine learning based computing method for extracting data from electronic documents, is disclosed. The machine learning based computing method includes steps of: receiving the electronic documents from data sources; converting formats of the electronic documents into first images associated with the electronic documents; classifying the first images associated with the electronic documents as second images including tabular data, and third images including non-tabular data, based on a CNN based machine learning model; determining tabular regions in tabular data associated with the second images by identifying bounding box coordinates indicating tabular regions in tabular data, based on a YOLO architecture based machine learning model; extracting first data from the tabular regions; extracting second data from non-tabular regions; and providing an output of the first data and the second data to second users on a user interface associated with electronic devices.

Abstract: A machine learning based computing method for generating electronic mails for accounts receivables management is disclosed.

Abstract: Systems, apparatus, articles of manufacture, and methods are disclosed. An example method includes: playing a first media stream in full screen; determining a user has selected a visual interface; after the user has selected the visual interface, presenting, in a first region of the visual interface, a group of tabs relevant to the user; after the user has selected the visual interface, playing, in a second region of the visual interface, the first media stream; determining the user has selected a second tab within the group; after the user has selected of the second tab, playing, in the second region of the visual interface, a second media stream corresponding to the second tab; after the user has selected of the second tab, determining the user has exited the visual interface without performing an additional selection; and after exiting the visual interface, playing the second media stream in full screen.

Abstract: The present invention is related to data processing methods and systems thereof. According to an embodiment, the present invention provides a method of processing documents using a machine learning model. The process begins by accessing data files and extracting information from them, which is subsequently stored. This document information, along with the machine learning model trained on various document formats, is used to classify the data files and generate tabular data. From this tabular data, data objects are created and included in an output data file. The information from the output file is then used to update the data of the machine learning model, optimizing it for improved future document processing. There are other embodiments as well.

Abstract: Generally discussed herein are devices, systems, and methods for globally accessible virtual private network (VPN) as a service. A method can include receiving a domain name request at a domain name system (DNS), the domain name request being received from a tenant premise equipment and including a domain name of a plurality of domain names associated with a VPN service, identifying, by the DNS and based on the domain name, an Internet Protocol (IP) address of a node of nodes dedicated to servicing traffic of the domain name, and establishing, by the DNS, a VPN tunnel between the tenant premise equipment and the identified node.

Abstract: Generally discussed herein are devices, systems, and methods for globally accessible virtual private network (VPN) as a service. A method can include receiving a domain name request at a domain name system (DNS), the domain name request being received from a tenant premise equipment and including a domain name of a plurality of domain names associated with a VPN service, identifying, by the DNS and based on the domain name, an Internet Protocol (IP) address of a node of nodes dedicated to servicing traffic of the domain name, and establishing, by the DNS, a VPN tunnel between the tenant premise equipment and the identified node.

Abstract: A network interface controller (NIC) associated with a virtual machine (VM) in a cloud computing network is configured to be flexibly attached and detached from a parent NIC to thereby enable the virtual machine to simultaneously be attached to multiple different virtual networks (VNets) and/or subnets that are associated with the same or different subscriptions. The inventive NIC, referred to herein as a flexibly extensible NIC (eNIC), enables a service provider to inject compute instances into an existing VNet using a multi-homing configuration in which the data plane uses a dedicated network interface to connect the customer's VNet, while another dedicated network interface provides management plane connectivity to the service provider. Such multi-VNet homing advantageously provides data plane isolation for the customer's VNet to comply with applicable security policies without disrupting management traffic between the injected resources and the service provider.