# US 2021/0279401 A1 · Combinatorial inflight analysis of multipart data > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 2021/0279401 A1 - Document ID: US20210279401A1 - Status: Published application - Kind: A1 - Application number: 17/326,812 - Filed: 2021-05-21 - Published: 2021-09-09 - Earliest claimed: 2020-02-28 - Assignees: Torch Research, LLC - Applicants: Torch Research, LLC - Inventors: Jon Kramer - Web page: https://patentgrep.com/patent/US20210279401A1 ## CPC - G06F 40/117: Tagging; Marking up (details of markup languages G06F40/143); Designating a block; Setting of attributes (style sheets, e.g. eXtensible Stylesheet Language Transformation [XSLT], G06F40/154) - G06F 40/30: Semantic analysis - G06F 16/254: {Extract, transform and load [ETL] procedures, e.g. ETL data flows in data warehouses} - G06F 16/9577: {Optimising the visualization of content, e.g. distillation of HTML documents} - G06F 40/106: Display of layout of documents; Previewing - G06F 40/295: Named entity recognition - G06N 3/0442: characterised by memory or gating, e.g. long short-term memory [LSTM] or gated recurrent units [GRU] - G06N 3/0464: Convolutional networks [CNN, ConvNet] - G06N 3/08: Learning methods - G06N 3/09: Supervised learning - G06N 20/00: Machine learning - G06F 40/143: Markup, e.g. Standard Generalized Markup Language [SGML] or Document Type Definition [DTD] - G06F 40/216: using statistical methods - G06N 3/044: Recurrent networks, e.g. Hopfield networks - G06N 3/045: Combinations of networks - G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N) - G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS ## Abstract A method of enhanced data orchestration (EDO). The method comprises receiving a first message by a mediation application executing on a computer system and analyzing the first message by the mediation application based at least in part on invoking a machine learning (ML) model by the mediation application. The analyzing determines that a feature of the first message is a probable first component of an item of personally identifiable information (PII). The method further comprises receiving a second message by the mediation application, determining by the mediation application that a feature of the second message when combined with the feature of the first message constitutes an item of PII, and treating the first message and the second message in accordance with predefined PII handling protocols by the mediation application. ## Claims ### Claim 1 A method of enhanced data orchestration (EDO), comprising: receiving a first message by a mediation application executing on a computer system; analyzing the first message by the mediation application based at least in part on invoking a machine learning (ML) model by the mediation application, wherein the analyzing determines that a feature of the first message is a probable first component of an item of personally identifiable information (PII); transmitting the first message to a receiving application; receiving a second message by the mediation application; determining by the mediation application that a feature of the second message when combined with the feature of the first message constitutes an item of PII; in response to the determining, attaching by the mediation application a tag to the second message identifying the feature of the second message and identifying the first message; and transmitting the second message with the tag attached to the receiving application, wherein the receiving application identifies the item of PII constituted by a combination of the first message and the second message while at least a portion of the item of PII is still in flight and not yet stored on a storage disk. ### Claim 2 (depends on 1) The method of claim 1, wherein the first message is held in non-permanent memory storage. ### Claim 3 (depends on 1) The method of claim 1, wherein the first message is held for a predetermined duration of time in non-permanent memory storage. ### Claim 4 (depends on 1) The method claim 1, wherein the mediation application further analyzes the first message based on one of an entity extraction ML model, a behavior anomaly ML model, a language translation ML model, or a sentiment analysis ML model. ### Claim 5 (depends on 4) The method of claim 4, wherein the mediation application attaches metadata to the first message, where the metadata is associated with results of the analysis associated with the one of the entity extraction ML model, the behavior anomaly ML model, the language translation ML model, or the sentiment analysis ML model. ### Claim 6 (depends on 2) The method of claim 2, further comprising attaching a second tag to the first message identifying the probable first component of PII and identifying the second message, wherein transmitting the first message comprises transmitting the first message with the second tag attached. ### Claim 7 A method of enhanced data orchestration (EDO), comprising: receiving a first message by a mediation application executing on a computer system; analyzing the first message by the mediation application based at least in part on invoking a machine learning (ML) model by the mediation application, wherein the analyzing determines that a feature of the first message is a probable first component of an item of personally identifiable information (PII); receiving a second message by the mediation application; determining by the mediation application that a feature of the second message when combined with the feature of the first message constitutes an item of PII; and treating the first message and the second message in accordance with predefined PII handling protocols by the mediation application. ### Claim 8 (depends on 7) The method of claim 7, wherein the first message is held in non-permanent memory storage. ### Claim 9 (depends on 7) The method of claim 7, wherein the first message is held for a predetermined duration of time in non-permanent memory storage. ### Claim 10 (depends on 7) The method claim 7, wherein the mediation application further analyzes the first message based on one of an entity extraction ML model, a behavior anomaly ML model, a language translation ML model, or a sentiment analysis ML model. ### Claim 11 (depends on 10) The method of claim 10, wherein the mediation application attaches metadata to the first message, where the metadata is associated with results of the analysis associated with the one of the entity extraction ML model, the behavior anomaly ML model, the language translation ML model, or the sentiment analysis ML model. ### Claim 12 (depends on 7) The method of claim 7, wherein treating the first message and the second message in accordance with predefined PII handling protocols comprises attaching a first tag to the first message identifying the probable first component of PII and identifying the second message and attaching a second tag to the second message identifying the feature of the second message and identifying the first message and transmitting the first message with the first tag attached and the second message with the second tag attached to a receiving application, whereby the receiving application is enabled to identify the item of PII constituted by a combination of the first message and the second message. ### Claim 13 (depends on 7) The method of claim 7, wherein the receiving application identifies the item of PII constituted by the combination of the first message and the second message while at least a portion of the item of PII is still in flight and not yet stored on a storage disk. ### Claim 14 A system for enhanced data orchestration (EDO), comprising: a processor; a non-transitory memory; and a medication application stored in the non-transitory memory, that when executed by the processor: receives a first message, analyzes the first message based at least in part on invoking a machine learning (ML) model by the mediation application, wherein the analyzing determines that a feature of the first message is a probable first component of an item of personally identifiable information (PII), receives a second message, determines that a feature of the second message when combined with the feature of the first message constitutes an item of PII, and treats the first message and the second message in accordance with predefined PII handling protocols by the mediation application. ### Claim 15 (depends on 14) The system of claim 14, wherein the first message is held in non-permanent memory storage. ### Claim 16 (depends on 14) The system of claim 14, wherein the first message is held for a predetermined duration of time in non-permanent memory storage. ### Claim 17 (depends on 14) The system claim 14, wherein the mediation application further analyzes the first message based on one of an entity extraction ML model, a behavior anomaly ML model, a language translation ML model, or a sentiment analysis ML model. ### Claim 18 (depends on 17) The system of claim 17, wherein the mediation application attaches metadata to the first message, where the metadata is associated with results of the analysis associated with the one of the entity extraction ML model, the behavior anomaly ML model, the language translation ML model, or the sentiment analysis ML model. ### Claim 19 (depends on 14) The system of claim 14, wherein treating the first message and the second message in accordance with predefined PII handling protocols comprises attaching a first tag to the first message identifying the probable first component of PII and identifying the second message and attaching a second tag to the second message identifying the feature of the second message and identifying the first message and transmitting the first message with the first tag attached and the second message with the second tag attached to a receiving application, whereby the receiving application is enabled to identify the item of PII constituted by a combination of the first message and the second message. ### Claim 20 (depends on 14) The system of claim 14, wherein the receiving application identifies the item of PII constituted by the combination of the first message and the second message while at least a portion of the item of PII is still in flight and not yet stored on a storage disk.