# US 12,118,019 B1 · Smart data signals for artificial intelligence based modeling > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 12,118,019 B1 - Document ID: US12118019B1 - Status: Granted - Kind: B1 - Application number: 18/523,876 - Filed: 2023-11-29 - Published: 2024-10-10 - Granted: 2024-10-15 - Earliest claimed: 2023-04-10 - Assignees: ExlService Holdings, Inc. - Applicants: ExlService Holdings, Inc. - Inventors: Lopamudra Panda; Sumit Taneja; Sumit Agarwal; Rashmi Ashrafi; Mustafa Karmalawala; Saurabh Mittal; Subodh Baranwal; Gregory Tyler Freeman; Shailesh Giri; Anurag Arora; Ajay Tiwari - Web page: https://patentgrep.com/patent/US12118019B1 ## CPC - G06F 16/285: {Clustering or classification} - G06F 3/0481: based on specific properties of the displayed interaction object or a metaphor-based environment, e.g. interaction with desktop elements like windows or icons, or assisted by a cursor's changing behaviour or appearance - G06F 3/04817: using icons (graphical or visual programming using iconic symbols G06F8/34) - G06F 16/24573: {using data annotations, e.g. user-defined metadata} - G06Q 40/08: Insurance - G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N) - G06Q: INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR ## Abstract A smart data signals platform for artificial intelligence/machine learning (AI/ML)-based modeling and simulation is structured to pre-process input data by structuring previously unstructured data that relates to a structured data item and linking the structured data to the data item to generate a validated enriched dataset. The validated enriched dataset is used to generate a trigger signal by evaluating at least an aspect to the enriched dataset against at least one criterion that corresponds to a loss hypothesis. The trigger signal is used to automatically monitor subsequently received structured data, access corresponding unstructured data and generate an analysis dataset for one or more machine learning models. The one or more machine learning models generate a computer-based prediction based on the analysis dataset. The prediction can include a resource degradation indicator. ## Claims ### Claim 1 A computer-implemented method executed by a smart data signals platform for artificial intelligence/machine learning (AI/ML)-based simulation and modeling, the method comprising: receiving, by a data enrichment engine from a first source computing system, first structured input data comprising claim data; receiving, by the data enrichment engine from a second source computing system, an unstructured input data comprising at least two of a call note, a claim note, an email, and a policy document; structuring the unstructured input data to generate second structured input data, wherein the second structured input data comprises a set of tokens generated based on the unstructured input data; linking, by the data enrichment engine, the first structured input data and the second structured input data to generate an enriched dataset comprising a set of metadata, wherein at least one item in the set of metadata corresponds to at least one token in the set of tokens; validating, by the data enrichment engine and based on a rule associated with a particular target computing system, the enriched dataset; performing, by a trigger engine, operations comprising: evaluating the enriched dataset against a criterion associated with a hypothesis; and based on a result of an evaluation, generating an electronic trigger signal; and deploying the electronic trigger signal in connection with the target computing system; performing, by a trained artificial intelligence (AI) decision engine, operations comprising: using the electronic trigger signal, generating an analysis dataset; executing a trained machine learning model on the analysis dataset to generate a resource degradation indicator; binding the resource degradation indicator to the analysis dataset; and transmitting the analysis dataset to the target computing system. ### Claim 2 (depends on 1) The method of claim 1, wherein the electronic trigger signal relates to claim progression through a sequence of events. ### Claim 3 (depends on 1) The method of claim 1, wherein the electronic trigger signal relates to a claim attribute. ### Claim 4 (depends on 1) The method of claim 1, wherein the electronic trigger signal is based on a previously generated signal. ### Claim 5 (depends on 1) The method of claim 1, wherein generating an analysis dataset comprises: generating, using the electronic trigger signal, the analysis dataset based on a subset of the enriched dataset; and enriching the analysis dataset with additional data received from a secondary computing system, the secondary computing system being at least one of an audit system, and email system, and a call logger system, the enriching comprising: generating and transmitting a data request to the secondary computing system, the data request comprising a claim identifier from the analysis dataset; and in response to receiving the additional data, binding the additional data to the analysis dataset as at least one of an additional data item or an additional metadata item. ### Claim 6 (depends on 5) The method of claim 5, wherein the additional data comprises at least one of a claim identifier and a claim attribute. ### Claim 7 (depends on 1) The method of claim 1, wherein the electronic trigger signal relates to a claim audit eligibility indicator, and wherein the claim audit eligibility indicator is determined according to a sampling criterion associated with the target computing system. ### Claim 8 (depends on 1) The method of claim 1, further comprising generating a graphical user interface (GUI) structured to display the analysis dataset. ### Claim 9 (depends on 8) The method of claim 8, wherein the GUI comprises a control structured to enable an operator to escalate a particular item based on the resource degradation indicator. ### Claim 10 (depends on 8) The method of claim 8, wherein the GUI comprises a control structured to display a first set of monitored claims and a second set of unmonitored claims from the claim data, the first set and the second set being determined based on the electronic trigger signal. ### Claim 11 (depends on 10) The method of claim 10, wherein the GUI comprises a control structured to enable an operator to drill down to a claim attribute, the claim attribute being at least one of a policy identifier, a claim identifier, a line of business, a priority, a due date, and a file review score. ### Claim 12 (depends on 10) The method of claim 10, wherein the GUI comprises a control structured to enable an operator to drill down to an electronically generated recommended action. ### Claim 13 (depends on 12) The method of claim 12, wherein the GUI comprises a control structured to enable an operator to override the electronically generated recommended action, and wherein an override is provided to the trained AI decision engine for model training. ### Claim 14 (depends on 13) The method of claim 13, wherein the GUI comprises a control structured to enable an operator to browse the first structured input data and the unstructured input data used to generate a recommended action, the method comprising: generating, based on the at least one item in the set of metadata, a hyperlink to a resource provided by the second source computing system to enable the operator to browse the unstructured input data. ### Claim 15 (depends on 1) The method of claim 1, wherein structuring the unstructured input data to generate second structured input data comprises: identifying a field in the unstructured input data; extracting from the field at least one item; and generating the set of tokens using the at least one item. ### Claim 16 (depends on 1) The method of claim 1, wherein the rule is managed, by a rules engine, for the particular target computing system or a set of target computing systems, and wherein the rule comprises an indication of a data value domain specific to the particular target computing system or the set of target computing systems. ### Claim 17 (depends on 16) The method of claim 16, wherein the data value domain relates to at least one of an incident type and a coverage type. ### Claim 18 (depends on 16) The method of claim 16, wherein validating the enriched dataset comprises updating or generating a token in the set of tokens to conform to a particular value or format in the data value domain. ### Claim 19 One or more non-transitory, computer-readable media having instructions stored thereon, the instructions, when executed by a processor of a smart data signals platform for artificial intelligence/machine learning (AI/ML)-based simulation and modeling, structured to cause the processor to perform operations comprising: receiving, by a data enrichment engine from a first source computing system, first structured input data comprising claim data; receiving, by the data enrichment engine from a second source computing system, an unstructured input data; structuring the unstructured input data to generate second structured input data, wherein the second structured input data comprises a set of tokens generated based on the unstructured input data; linking, by the data enrichment engine, the first structured input data and the second structured input data to generate an enriched dataset comprising a set of metadata, wherein at least one item in the set of metadata corresponds to at least one token in the set of tokens; validating, by the data enrichment engine and based on a rule associated with a particular target computing system, the enriched dataset; performing, by a trigger engine, operations comprising: evaluating the enriched dataset against a criterion associated with a hypothesis; and based on a result of an evaluation, generating an electronic trigger signal; and deploying the electronic trigger signal in connection with the target computing system; performing, by a trained artificial intelligence (AI) decision engine, operations comprising: using the electronic trigger signal, generating an analysis dataset; executing a trained machine learning model on the analysis dataset to generate a resource degradation indicator; binding the resource degradation indicator to the analysis dataset; and transmitting the analysis dataset to the target computing system. ### Claim 20 A computing system comprising at least one processor, at least one memory, and one or more non-transitory, computer-readable media having instructions stored thereon, the instructions, when executed by the at least one processor, causing a smart data signals platform to perform operations for artificial intelligence/machine learning (AI/ML)-based simulation and modeling, the operations comprising: receive, by a data enrichment engine from a first source computing system, first structured input data comprising claim data; receive, by the data enrichment engine from a second source computing system, an unstructured input data; structure the unstructured input data to generate second structured input data, wherein the second structured input data comprises a set of tokens generated based on the unstructured input data; link, by the data enrichment engine, the first structured input data and the second structured input data to generate an enriched dataset comprising a set of metadata, wherein at least one item in the set of metadata corresponds to at least one token in the set of tokens; validate, by the data enrichment engine and based on a rule associated with a particular target computing system, the enriched dataset; perform, by a trigger engine, operations comprising: evaluating the enriched dataset against a criterion associated with a hypothesis; and based on a result of an evaluation, generating an electronic trigger signal; and deploying the electronic trigger signal in connection with the target computing system; perform, by a trained artificial intelligence (AI) decision engine, operations comprising: using the electronic trigger signal, generating an analysis dataset; executing a trained machine learning model on the analysis dataset to generate a resource degradation indicator; bind the resource degradation indicator to the analysis dataset; and transmit the analysis dataset to the target computing system.