# US 2023/0274282 A1 · Transaction tracking and fraud detection using voice and/or video data > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 2023/0274282 A1 - Document ID: US20230274282A1 - Status: Published application - Kind: A1 - Application number: 18/168,458 - Filed: 2023-02-13 - Published: 2023-08-31 - Earliest claimed: 2020-01-22 - Assignees: Capital One Services, LLC - Applicants: Capital One Services, LLC - Inventors: Abdelkader M'Hamed BENKREIRA; Joshua EDWARDS; Michael MOSSOBA - Web page: https://patentgrep.com/patent/US20230274282A1 ## CPC - G06Q 20/4016: {involving fraud or risk level assessment in transaction processing} - G06N 20/00: Machine learning - 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 - G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS ## Abstract A device receives recording data, for a recording of a user associated with an account, that captures the user describing a transaction. The device processes the recording data to identify one or more characteristics of an individual that described the transaction in the recording. The device determines, based on the one or more characteristics of the individual, whether the individual that described the transaction in the recording is the user associated with the account. The device causes the recording data to be stored in association with transaction data that identifies a list of transactions that are associated with the account of the user, wherein causing the recording data to be stored in association with the transaction data allows the recording to be made accessible to the user via an interface of an application used to manage the account. ## Claims ### Claim at position 1 1-20. (canceled) ### Claim 21 A method, comprising: analyzing, by a device and based on receiving an indication that a transaction involving an account is to be processed, recording data to determine one or more transaction characteristics that describe the transaction, wherein the recording data includes a recording of an authorized user associated with the account describing the transaction; applying, by the device, the one or more transaction characteristics as input to a risk score model; determining, by the device, a risk score associated with the transaction based on an output of the risk score model, wherein the risk score model is a neural network model trained on historical transaction data including historical transaction characteristics; and causing, by the device and based on the risk score meeting or exceeding a threshold, the recording data to be stored in association with transaction data that identifies a list of transactions that are associated with the account. ### Claim 22 (depends on 1) The method of claim 1, wherein causing the recording data to be stored comprises: causing a data structure to store the recording to enable the authorized user to search for the recording via a search feature. ### Claim 23 (depends on 1) The method of claim 1, wherein the recording captures one or more gestures associated with the authorized user. ### Claim 24 (depends on 1) The method of claim 1, further comprising: comparing a first dataset that identifies a set of facial characteristics of an individual, associated with the recording, with a second dataset that identifies a corresponding set of facial characteristics of the authorized user, or comparing a third dataset that identifies the set of voice-related characteristics of the individual with a fourth dataset that identifies a corresponding set of voice-related characteristics of the authorized user. ### Claim 25 (depends on 1) The method of claim 1, further comprising: using machine learning to determine whether an individual associated with the recording is the authorized user. ### Claim 26 (depends on 1) The method of claim 1, further comprising: identifying one or more missing transaction characteristics that are not identified in the recording; and performing one or more actions based on identifying the one or more missing transaction characteristics. ### Claim 27 (depends on 1) The method of claim 1, further comprising: determining to permit the transaction or prevent the transaction from being processed based on the risk score. ### Claim 28 A device, comprising: one or more memories; and one or more processors, coupled to the one or more memories, configured to: analyze, based on receiving an indication that a transaction involving an account is to be processed, recording data to determine one or more transaction characteristics that describe the transaction, wherein the recording data includes a recording of an authorized user associated with the account describing the transaction; apply the one or more transaction characteristics as input to a risk score model; determine a risk score associated with the transaction based on an output of the risk score model, wherein the risk score model is a neural network model trained on historical transaction data including historical transaction characteristics; and cause, based on the risk score meeting or exceeding a threshold, the recording data to be stored in association with transaction data that identifies a list of transactions that are associated with the account. ### Claim 29 (depends on 8) The device of claim 8, wherein the one or more processors, to cause the recording data to be stored, are configured to: cause a data structure to store the recording to enable the authorized user to search for the recording via a search feature. ### Claim 30 (depends on 8) The device of claim 8, wherein the recording captures one or more gestures associated the authorized user. ### Claim 31 (depends on 8) The device of claim 8, wherein the one or more processors are further configured to: compare a first dataset that identifies a set of facial characteristics of an individual, associated with the recording, with a second dataset that identifies a corresponding set of facial characteristics of the authorized user, or compare a third dataset that identifies the set of voice-related characteristics of the individual with a fourth dataset that identifies a corresponding set of voice-related characteristics of the authorized user. ### Claim 32 (depends on 8) The device of claim 8, wherein the one or more processors are further configured to: use machine learning to determine whether an individual associated with transaction is the authorized user. ### Claim 33 (depends on 8) The device of claim 8, wherein the one or more processors are further configured to: identify one or more missing transaction characteristics that are not identified in the recording; and perform one or more actions based on identifying the one or more missing transaction characteristics. ### Claim 34 (depends on 8) The device of claim 8, wherein the one or more processors are further configured to: determine to permit the transaction or prevent the transaction from being processed based on the risk score. ### Claim 35 A non-transitory computer-readable medium storing a set of instructions, the set of instructions comprising: one or more instructions that, when executed by one or more processors of a device, cause the device to: analyze, based on receiving an indication that a transaction involving an account is to be processed, recording data to determine one or more transaction characteristics that describe the transaction, wherein the recording data includes a recording of an authorized user associated with the account describing the transaction; apply the one or more transaction characteristics as input to a risk score model; determine a risk score associated with the transaction based on an output of the risk score model, wherein the risk score model is a neural network model trained on historical transaction data including historical transaction characteristics; and cause, based on the risk score meeting or exceeding a threshold, the recording data to be stored in association with transaction data that identifies a list of transactions that are associated with the account. ### Claim 36 (depends on 15) The non-transitory computer-readable medium of claim 15, wherein the one or more instructions, that cause the device to cause the recording data to be stored, cause the device to: cause a data structure to store the recording to enable the authorized user to search for the recording via a search feature. ### Claim 37 (depends on 15) The non-transitory computer-readable medium of claim 15, wherein the one or more instructions further cause the device to: compare a first dataset that identifies a set of facial characteristics of an individual, associated with the recording, with a second dataset that identifies a corresponding set of facial characteristics of the authorized user, or compare a third dataset that identifies the set of voice-related characteristics of the individual with a fourth dataset that identifies a corresponding set of voice-related characteristics of the authorized user. ### Claim 38 (depends on 15) The non-transitory computer-readable medium of claim 15, wherein the one or more instructions further cause the device to: use machine learning to determine whether an individual associated with transaction is the authorized user. ### Claim 39 (depends on 15) The non-transitory computer-readable medium of claim 15, wherein the one or more instructions further cause the device to: identify one or more missing transaction characteristics that are not identified in the recording; and perform one or more actions based on identifying the one or more missing transaction characteristics. ### Claim 40 (depends on 15) The non-transitory computer-readable medium of claim 15, wherein the one or more instructions further cause the device to: determine to permit the transaction or prevent the transaction from being processed based on the risk score.