IP Library › Granted Patent US 12,737,814
Granted Patent B2
US 12,737,814 · App. 19/534,069 · Granted Sep 15, 2026

Systems and methods for automatically classifying distributed user-state data across a plurality of heterogeneous data sources

Inventors: Jessica Willis (Huntington Woods, MI); Chris Wascha (Detroit, MI); Erik Gratz (Souix Falls, SD); Joel Powell (Detroit, MI); Tucker Steelman (Tamala, CA)
Assignee: PocketNest, Inc.
G06Q40/06G06Q10/06311G06Q40/02
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 12,737,814
App. No.
19/534,069
Granted
Sep 15, 2026
Kind
B2
Abstract

A system and method for automatically classifying distributed user state data across a plurality of heterogeneous data sources may include receiving encrypted user state data, generating a normalized user-state data structure based on the encrypted user state data, generating a completion-state classification vector based on the normalized user-state data structure, and automatically propagating the normalized user-state data and/or completion-state data to one or more third-party computing services.

Claims (107)

1 . A computer-implemented system for classifying distributed user state data across a plurality of heterogeneous data sources, the computer-implemented system comprising:

a memory storing a plurality of computer-executable instructions and maintaining:

a normalized user-state data structure that comprises normalized, machine-readable user state data converted from a plurality of heterogeneous data sources;

a completion-state propagation index that associates one or more completion-state classification vectors with one or more service identifiers of one or more third-party computing services; and

one or more processors that execute the plurality of computer-executable instructions to perform operations comprising:

receiving, from the plurality of heterogeneous data sources, encrypted user state data collected through a sequence of event-driven user interfaces, wherein the encrypted user state data includes encrypted user-entered responses and encrypted account data obtained from one or more linked external accounts;

decrypting, by the one or more processors executing a computer decryption program, the encrypted user state data into decrypted user state data and transforming the decrypted user state data into the normalized user-state data structure by:

extracting, by a feature extractor, a corpus of raw user state features from the decrypted user state data,

converting, by the one or more processors, the corpus of raw user state features into normalized user state features associated with the normalized user-state data structure, and

embedding the normalized user state features into the normalized user-state data structure;

generating, based on the normalized user-state data structure, a completion-state classification vector of the one or more completion-state classification vectors comprising a completion state classification value for a plurality of state-dependent tasks;

in response to generating the completion-state classification vector:

generating, by the one or more processors, one or more positive state-completion signals indicating that the completion state classification value for a subset of the plurality of state-dependent tasks satisfy predefined state-completion criteria;

generating, by the one or more processors, one or more negative state-completion signals indicating that the completion state classification value for a remainder of the plurality of state-dependent tasks do not satisfy the predefined state-completion criteria; and

generating a real-time interactive visualization of the completion-state classification vector and the one or more positive and negative state-completion signals;

requesting, from an authorized user associated with the encrypted user state data, permission to display the real-time interactive visualization of the completion-state classification vector and the one or more positive and negative state-completion signals;

transmitting, via the one or more processors, the real-time interactive visualization to a device of the authorized user after the permission is granted; and

automatically providing, via the one or more processors, the one or more third-party computing services with access to (i) the completion-state classification vector generated for the authorized user and (ii) the real-time interactive visualization of the completion-state classification vector, and (iii) the normalized user-state data structure in response to the authorized user subscribing to each of the one or more third-party computing services, wherein the completion-state propagation index is updated to store a machine-readable association between the completion-state classification vector of the authorized user and each of the one or more third-party computing services, wherein in response to the authorized user unsubscribing from a respective third-party computing service:

the one or more processors remove, from the completion-state propagation index, the machine-readable association between the completion-state classification vector of the authorized user and each of the one or more third-party computing services, thereby preventing subsequent access by the respective third-party computing service.

2 . The computer-implemented system according to claim 1 , wherein generating the completion state classification value for a first respective state-dependent task of the plurality of state-dependent tasks includes:

extracting, from the normalized user-state data structure, a first plurality of user state columns mapped to the first respective state-dependent task;

in response to extracting the first plurality of user state columns mapped to the first respective state-dependent task, generating a first binary state-bit value encoding that each of the first plurality of user state columns comprises a non-null value; and

embedding, into the completion-state classification vector, the first binary state-bit value at a first index position of the completion-state classification vector corresponding to the first respective state-dependent task.

3 . The computer-implemented system according to claim 2 , wherein generating the completion state classification value for a second respective state-dependent task of the plurality of state-dependent tasks includes:

extracting, from the normalized user-state data structure, a second plurality of user state columns mapped to the second respective state-dependent task, different from the first plurality of user state columns;

in response to extracting the second plurality of user state columns mapped to the second respective state-dependent task, generating a second binary state-bit value encoding that at least one the second plurality of user state columns comprises a null value; and

embedding, into the completion-state classification vector, the second binary state-bit value at a second index position of the completion-state classification vector corresponding to the second respective state-dependent task.

4 . The computer-implemented system according to claim 1 , wherein generating the one or more positive and negative state-completion signals and the real-time interactive visualization includes:

detecting, by the one or more processors, that a first index position of the completion-state classification vector includes a first binary state-bit value encoded to indicate that at least one of a first plurality of user state columns in the normalized user-state data structure comprise a null value;

in response to detecting the first index position of the completion-state classification vector, broadcasting, to a primary user interface controller node, a first real-time broadcast message comprising a first key-value payload comprising:

a first key-value pair comprising a state-dependent task key and a first respective state-dependent task corresponding to the first index position as a value of the state-dependent task key, and

a second key-value pair comprising a completion state classification value key and the first binary state-bit value as a value of the completion state classification value key;

creating, in the real-time interactive visualization, a first visual task element based on the value of the completion state classification value key in the first real-time broadcast message; and

setting, by the primary user interface controller node, the first visual task element to a first chromatic intensity based on the value of the state-dependent task key in the first real-time broadcast message.

5 . The computer-implemented system according to claim 4 , wherein the one or more processors that execute the plurality of computer-executable instructions perform the operations comprising:

generating, via the sequence of event-driven user interfaces, updated encrypted user state data for the first respective state-dependent task;

in response to generating the updated encrypted user state data for the first respective state-dependent task:

detecting, by the one or more processors, that the first index position of the completion-state classification vector has been updated from the first binary state-bit value to a second binary state-bit value encoded to indicate that each of the first plurality of user state columns in the normalized user-state data structure comprises a non-null value;

broadcasting, in response to detecting that the first index position has been updated to the second binary state-bit value, an update to the first real-time broadcast message that changes the value of the completion state classification value key from the first binary state-bit value to the second binary state-bit value; and

dynamically changing, by the primary user interface controller node in real-time, the first visual task element from the first chromatic intensity to a second chromatic intensity in response to broadcasting the update to the first real-time broadcast message.

6 . The computer-implemented system according to claim 4 , wherein generating the one or more positive and negative state-completion signals and the real-time interactive visualization includes:

detecting, by the one or more processors, that a second index position of the completion-state classification vector includes a second binary state-bit value encoded to indicate that each of a second plurality of user state columns in the normalized user-state data structure comprise a non-null value;

in response to detecting the second index position of the completion-state classification vector, broadcasting, to the primary user interface controller node, a second real-time broadcast message comprising a second key-value payload comprising:

a third key-value pair comprising the state-dependent task key and a second respective state-dependent task corresponding to the second index position as the value of the state-dependent task key, and

a fourth key-value pair comprising the completion state classification value key and the second binary state-bit value as the value of the completion state classification value key;

creating, in the real-time interactive visualization, a second visual task element based on the value of the completion state classification value key in the second real-time broadcast message; and

setting, by the primary user interface controller node, the second visual task element to a second chromatic intensity based on the value of the state-dependent task key in the second real-time broadcast message.

7 . The computer-implemented system according to claim 1 , wherein requesting the permission to display the real-time interactive visualization and transmitting the real-time interactive visualization to the device of the authorized user includes:

transmitting, via the one or more processors, an electronic communication to the authorized user comprising:

a first graphical user interface button that grants the permission to display the real-time interactive visualization on the device of the authorized user, and

a second graphical user interface button that denies the permission to display the real-time interactive visualization on the device of the authorized user;

automatically installing, on the device of the authorized user, an application bundle in response to receiving a selection of the first graphical user interface button; and

automatically launching the application bundle and navigating to the real-time interactive visualization after automatically installing the application bundle on the device of the authorized user.

8 . The computer-implemented system according to claim 7 , wherein the one or more processors that execute the plurality of computer-executable instructions perform the operations comprising:

after automatically providing the one or more third-party computing services with access to (i) the completion-state classification vector generated for the authorized user, (ii) the real-time interactive visualization of the completion-state classification vector, and (iii) the normalized user-state data structure in response to the authorized user subscribing to each of the one or more third-party computing services:

detecting, by the application bundle, that a third-party application associated with the one or more third-party computing services is open at the device of the authorized user; and

establishing a bi-directional communication channel between the third-party application and the one or more processors in response to detecting that the third-party application is open;

automatically transmitting, via the bi-directional communication channel, the one or more positive state-completion signals to the third-party application; and

in response to the third-party application receiving the one or more positive state-completion signals, automatically inhibiting execution of an activated function of the third-party application corresponding to the subset of the plurality of state-dependent tasks that satisfy the pre-defined state-completion criteria.

9 . The computer-implemented system according to claim 8 , wherein the one or more processors that execute the plurality of computer-executable instructions perform the operations comprising:

after automatically providing the one or more third-party computing services with access to (i) the completion-state classification vector generated for the authorized user, (ii) the real-time interactive visualization of the completion-state classification vector, and (iii) the normalized user-state data structure in response to the authorized user subscribing to each of the one or more third-party computing services:

automatically transmitting, via the bi-directional communication channel, the one or more negative state-completion signals to the third-party application; and

automatically surfacing, via a popup window of the third-party application, one or more unused or unavailable functions of the third-party application associated with each of the remainder of the plurality of state-dependent tasks that do not satisfy the predefined state-completion criteria.

10 . The computer-implemented system according to claim 7 , wherein the one or more processors that execute the plurality of computer-executable instructions perform the operations comprising:

after automatically providing the one or more third-party computing services with access to (i) the completion-state classification vector generated for the authorized user, (ii) the real-time interactive visualization of the completion-state classification vector, and (iii) the normalized user-state data structure in response to the authorized user subscribing to each of the one or more third-party computing services:

detecting, by the application bundle, that a third-party application associated with the one or more third-party computing services is open at the device of the authorized user; and

establishing, via the one or more processors, a bi-directional communication channel between the third-party application and the one or more processors in response to detecting that the third-party application is open;

automatically transmitting, via the bi-directional communication channel, the one or more positive state-completion signals to the third-party application; and

in response to the third-party application receiving the one or more positive state-completion signals, automatically providing, via the third-party application, a third-party system-generated credential that was not previously available to the authorized user before the one or more positive state-completion signals were transmitted.

11 . The computer-implemented system according to claim 1 , wherein converting the corpus of raw user state features into the normalized user state features includes:

extracting, via the one or more processors, the normalized user state features associated with the normalized user-state data structure;

constructing, via the one or more processors, a machine-readable classification prompt that specifies a normalization schema for mapping the corpus of raw user state features to the normalized user state features;

inputting, via the one or more processors, the machine-readable classification prompt to a large language model; and

receiving, in response to the large language model executing the machine-readable classification prompt, a machine-readable output that maps each raw user state feature in the corpus of raw user state features to a respective normalized user state feature of the normalized user state features.

12 . The computer-implemented system according to claim 11 , wherein the normalization schema for mapping the corpus of raw user state features to the normalized user state features comprises:

a plurality of raw input feature identifiers corresponding to the corpus of raw user state features;

a plurality of normalized feature identifiers corresponding to the normalized user state features;

a required output structure defining a set of parameterized schema elements, each parameterized schema element of the set of parameterized schema elements comprising:

a respective normalized feature identifier corresponding to a distinct one of the plurality of normalized feature identifiers; and

a placeholder parameter that prompts the large language model to replace text associated with the placeholder parameter with a raw input feature identifier corresponding to the respective normalized feature identifier.

13 . The computer-implemented system according to claim 12 , wherein the one or more processors that execute the plurality of computer-executable instructions perform the operations comprising:

detecting, by the one or more processors, that the text of the placeholder parameter has not been replaced with the raw input feature identifier; and

inserting, in place of the placeholder parameter, a null identifier corresponding to the respective normalized feature identifier.

14 . The computer-implemented system according to claim 1 , wherein the one or more processors that execute the plurality of computer-executable instructions perform the operations comprising:

detecting, in real-time, a change in a completion state classification value at a corresponding index position of a respective completion-state classification vector of the one or more completion-state classification vectors stored in the completion-state propagation index; and

propagating the change to each third-party computing service associated with the completion-state classification vector by transmitting only an index position identifier and an updated binary state-bit value, without regenerating the completion-state classification vector or reprocessing the normalized user-state data structure.

15 . A computer-implemented method for classifying distributed user state data across a plurality of heterogeneous data sources, the computer-implemented method comprising:

receiving, from the plurality of heterogeneous data sources, encrypted user state data collected through a sequence of event-driven user interfaces, wherein the encrypted user state data includes encrypted user-entered responses and encrypted account data obtained from one or more linked external accounts;

decrypting, by the one or more processors executing a computer decryption program, the encrypted user state data into decrypted user state data and transforming the decrypted user state data into a normalized user-state data structure by:

extracting, by a feature extractors, a corpus of raw user state features from the decrypted user state data,

converting, by the one or more processors, the corpus of raw user state features into normalized user state features associated with the normalized user-state data structure, and

embedding the normalized user state features into the normalized user-state data structure;

generating, based on the normalized user-state data structure, a completion-state classification vector of the one or more completion-state classification vectors comprising a completion state classification value for a plurality of state-dependent tasks, wherein generating the completion state classification value for a first respective state-dependent task of the plurality of state-dependent tasks includes:

extracting, from the normalized user-state data structure, a first plurality of user state columns mapped to the first respective state-dependent task,

in response to extracting the first plurality of user state columns mapped to the first respective state-dependent task, generating a first binary state-bit value encoding that each of the first plurality of user state columns comprises a non-null value, and

embedding, into the completion-state classification vector, the first binary state-bit value at a first index position of the completion-state classification vector corresponding to the first respective state-dependent task;

in response to generating the completion-state classification vector:

generating, by the one or more processors, one or more positive state-completion signals indicating that the completion state classification value for a subset of the plurality of state-dependent tasks satisfy predefined state-completion criteria;

generating, by the one or more processors, one or more negative state-completion signals indicating that the completion state classification value for a remainder of the plurality of state-dependent tasks do not satisfy the predefined state-completion criteria; and

generating a real-time interactive visualization of the completion-state classification vector and the one or more positive and negative state-completion signals;

requesting, from an authorized user associated with the encrypted user state data, permission to display the real-time interactive visualization of the completion-state classification vector and the one or more positive and negative state-completion signals;

transmitting, via the one or more processors, the real-time interactive visualization to a device of the authorized user after the permission is granted; and

automatically providing, via the one or more processors, the one or more third-party computing services with access to (i) the completion-state classification vector generated for the authorized user and (ii) the real-time interactive visualization of the completion-state classification vector, and (iii) the normalized user-state data structure in response to the authorized user subscribing to each of the one or more third-party computing services, wherein a completion-state propagation index is updated to store a machine-readable association between the completion-state classification vector of the authorized user and each of the one or more third-party computing services.

16 . The computer-implemented method according to claim 15 , wherein generating the completion state classification value for a second respective state-dependent task of the plurality of state-dependent tasks includes:

extracting, from the normalized user-state data structure, a second plurality of user state columns mapped to the second respective state-dependent task, different from the first plurality of user state columns;

in response to extracting the second plurality of user state columns mapped to the second respective state-dependent task, generating a second binary state-bit value encoding that at least one the second plurality of user state columns comprises a null value; and

embedding, into the completion-state classification vector, the second binary state-bit value at a second index position of the completion-state classification vector corresponding to the second respective state-dependent task.

Continuity (4)
Continuation In Part 18796675 · Aug 7, 2024
Continuation 16864599 · May 1, 2020
Provisional Application 62842256 · May 2, 2019
Related Publication 20260203820A1 · Jul 16, 2026
References Cited (14)
US 8200573B2 · Fineout · 2012 [cited by examiner]
US 8548903B2 · Becker · 2013 [cited by examiner]
US 10664247B1 · Hubbard · 2020 [cited by examiner]
US 11681811B1 · Dixit · 2023 [cited by examiner]
US 11853982B1 · Woodward · 2023 [cited by examiner]
US 20080228640A1 · Brown · 2008 [cited by examiner]
US 20090106141A1 · Becker · 2009 [cited by examiner]
US 20100332355A1 · Lopez · 2010 [cited by examiner]
US 20180075535A1 · Alvarenga · 2018 [cited by examiner]
US 20230130840A1 · Olesen · 2023 [cited by examiner]
US 20250335998A1 · Hurwitz · 2025 [cited by examiner]
US 20260134474A1 · Patton · 2026 [cited by examiner]
Vitor Valerio de Souza Campos et al., “Review and comparison of works on heterogeneous data and semantic analysis in Big Data”, Semina: Ciênc. Ex. Tech., Londrina, v. 42, n. 1, p. 113-128, Jan./Jun. (Year: 2021). [cited by examiner]
Putrama et al., Heterogeneous data integration: Challenges and Opportunities, a Department of Electronics Technology, Faculty of Electrical Engineering and Informatics, Budapest University of Technology and Economics, B… [cited by examiner]