Granted Patent B2

US 12,367,971 · App. 18/900,253 · Granted Jul 22, 2025

Inventors: Munjal Shah (Palo Alto, CA); Vishal Parikh (Palo Alto, CA); Meenesh Bhimani (Palo Alto, CA); Subhabrata Mukherjee (Palo Alto, CA); Alex Miller (Palo Alto, CA); Saad Godil (Palo Alto, CA); Debajyoti Datta (Palo Alto, CA); Paul Gamble (Palo Alto, CA); Rae Lasko (Palo Alto, CA)

Assignee: HealthGPT, Inc.

G16H40/20G10L15/183G10L15/22G10L25/66G10L15/1822G10L15/19

Abstract

Claims (87)

1. A method, comprising:

initiating, by an artificial intelligence executed by one or more processors, a conversation with a human, the artificial intelligence configured to complete a checklist during the conversation;

receiving, by the artificial intelligence, a first human response from the human;

extracting, by the one or more processors, data from the first human response;

storing, by the one or more processors, the data in a conversation summary;

providing, by the artificial intelligence, a first artificial intelligence response to the human;

generating, by the artificial intelligence and during the conversation with the human, multiple possible responses based at least in part on the data in the conversation summary;

receiving, by the artificial intelligence, a second human response from the human;

accessing, by the artificial intelligence, the conversation summary;

providing to the human, by the artificial intelligence, a second artificial intelligence response selected from the multiple possible responses based at least in part on a portion of the data in the conversation summary; and

based at least in part on determining, by the artificial intelligence, that the checklist has been completed, ending the conversation with the human.

2. The method of claim 1 , wherein the conversation summary is cached to reduce a time to access the conversation summary.

3. The method of claim 1 , wherein the conversation summary comprises a knowledge graph.

4. The method of claim 1 , wherein accessing the conversation summary comprises:

determining that the second human response references information previously provided by the human; and

identifying, in the conversation summary, the information previously provided by the human.

5. The method of claim 1 , wherein after receiving, by the artificial intelligence, the first human response from the human, the method comprises:

providing the first human response to a second artificial intelligence to determine additional factors associated with the human in parallel with providing, by the artificial intelligence, the first artificial intelligence response to the human.

6. The method of claim 1 , wherein the artificial intelligence is trained using multi-turn reinforcement learning through human feedback (RLHF).

7. The method of claim 1 , wherein the artificial intelligence, during the conversation with the human:

identifies a turn-yielding cue;

performs interruption detection to detect when the human is attempting to interrupt the artificial intelligence;

identifies a non-verbal cue associated with the human; or

any combination thereof.

8. A server comprising:

one or more processors; and

one or more computer-readable storage media to store instructions executable by the one or more processors to perform operations comprising:

initiating, by an artificial intelligence, a conversation with a human, the artificial intelligence configured to complete a checklist during the conversation;

receiving, by the artificial intelligence, a first human response from the human;

extracting data from the first human response;

storing the data in a conversation summary;

providing, by the artificial intelligence, a first artificial intelligence response to the human;

generating, by the artificial intelligence and during the conversation with the human, multiple possible responses based at least in part on the data in the conversation summary;

receiving, by the artificial intelligence, a second human response from the human;

accessing, by the artificial intelligence, the conversation summary;

providing to the human, by the artificial intelligence, a second artificial intelligence response selected from the multiple possible responses based at least in part on a portion of the data in the conversation summary; and

based at least in part on determining, by the artificial intelligence, that the checklist has been completed, ending the conversation with the human.

9. The server of claim 8 , wherein the artificial intelligence is configured to perform specialized healthcare-related functions comprising one or more of:

gathering data related to performing Healthcare Effectiveness Data and Information Set (HEDIS) calculations;

performing a Health Records Assessment (HRA);

determining a Risk Adjustment Factor (RAF);

reviewing a pre-op checklist;

reviewing a discharge checklist;

reviewing a chronic care checklist;

determining social determinants of health (SDOH); or

any combination thereof.

10. The server of claim 8 , wherein the artificial intelligence comprises:

a task completion engine;

an automatic speech recognition (ASR) module;

a speech synthesis module;

a retrieval augmented generation (RAG) module; and

a conversation engine.

11. The server of claim 10 , wherein the conversation engine comprises:

a turn engine;

a dialect detection module;

a predictive answering module;

a tone detection module; and

an interruption detection module.

12. The server of claim 8 , wherein accessing the conversation summary comprises:

determining that the second human response references information previously provided by the human; and

identifying, in the conversation summary, the information previously provided by the human.

13. The server of claim 8 , wherein the conversation summary is stored in a cache.

14. The server of claim 8 , wherein after receiving, by the artificial intelligence, the first human response from the human, the operations comprise:

providing the first human response to a second artificial intelligence to determine additional factors associated with the human in parallel with providing, by the artificial intelligence, the first artificial intelligence response to the human.

15. A non-transitory memory device to store instructions executable by one or more processors to perform operations comprising:

initiating, by an artificial intelligence, a conversation with a human, the artificial intelligence configured to complete a checklist during the conversation;

receiving, by the artificial intelligence, a first human response from the human;

extracting data from the first human response;

storing the data in a conversation summary;

providing, by the artificial intelligence, a first artificial intelligence response to the human;

generating, by the artificial intelligence and during the conversation with the human, multiple possible responses based at least in part on the data in the conversation summary;

receiving, by the artificial intelligence, a second human response from the human;

accessing, by the artificial intelligence, the conversation summary;

providing to the human, by the artificial intelligence, a second artificial intelligence response selected from the multiple possible responses based at least in part on a portion of the data in the conversation summary; and

based at least in part on determining, by the artificial intelligence, that the checklist has been completed, ending the conversation with the human.

16. The non-transitory memory device of claim 15 , wherein the artificial intelligence engages in a task that includes one or more of:

a preventative screening;

a intake-related task;

a scheduling-related task;

a pre-op related tasks;

a discharge-related task;

a chronic care related task; or

any combination thereof.

17. The non-transitory memory device of claim 16 , wherein the discharge-related task includes an explanation of benefits and billing.

18. The non-transitory memory device of claim 15 , wherein the conversation summary is cached.

19. The non-transitory memory device of claim 15 , wherein at least a portion of the conversation is text-based.

20. The non-transitory memory device of claim 15 , wherein at least a portion of the conversation is audio-based.

Assignments (1)

Continuity (3)

References Cited (12)