Back to results
US 12,762,139 B1Granted
Artificial intelligence agent execution using triggers
AssigneeNotion Labs, Inc.
InventorsQuinn Hajime Johnson, Sean Keenan, Eric Liu, Ryan Nystrom, Simon Townsend-Last
Life of this patent
Estimating the term…
Abstract
Disclosed herein are systems, methods, and computer readable media for implementing artificial intelligence (AI) agents that are configured with automations. In some embodiments, an automation is configured for a first AI agent, with the automation including a trigger that automatically initiates a call of the first AI agent. An occurrence of the trigger may be detected and, in response, the first AI agent may be called and provided with context associated with the trigger. An output resulting from execution of the first set of instructions by the first AI agent may be received.
Claims18
One or more non-transitory, computer-readable storage media comprising instructions recorded thereon, wherein the instructions, when executed by at least one data processor of a system, cause the system to:
- configure an automation for a first artificial intelligence (AI) agent,
- wherein the automation includes a trigger that automatically initiates a call of the first AI agent,
- wherein calling the first AI agent causes the first AI agent to execute a first set of instructions, the first set of instructions including an instruction to execute a tool for modifying a database within a workspace application, and
- wherein the trigger is an action performed within either the workspace application or a third-party application;
- detect an occurrence of the trigger;
- in response to detecting the occurrence of the trigger, call the first AI agent and provide context associated with the trigger to the first AI agent;
- receive, from the first AI agent, an output resulting from execution of the first set of instructions by the first AI agent,
- wherein the first AI agent generates the output by:
- (i) determining a semantic meaning of the context associated with the trigger, (ii) comparing the semantic meaning to one or more descriptions associated with one or more tools from a set of tools, (iii) based on the comparing, determining a list of usable tools from the set of tools for modifying the database, (iv) based on the list of usable tools and the semantic meaning, generating a first plan including the first set of instructions, and (v) executing one or more instructions from the first set of instructions, and
- wherein the output includes a modification to the database;
- direct a second AI agent to determine whether the output indicates a successful execution of the first set of instructions;
- upon receiving an indication from the second AI agent that the output is an unsuccessful execution of the first set of instructions:
- cause display, via a user interface, of a notification of unsuccessful execution;
- and
- upon receiving an indication from the second AI agent that the output is a successful execution of the first set of instructions:
- cause display, via the user interface, of the output.
13 more elements
A system comprising:
- at least one hardware processor;
- and
- at least one non-transitory memory storing instructions, which, when executed by the at least one hardware processor, cause the system to:
- configure an automation for a first artificial intelligence (AI) agent,
- wherein the automation includes a trigger that automatically initiates a call of the first AI agent, and
- wherein calling the first AI agent causes the first AI agent to execute a first set of instructions;
- detect an occurrence of the trigger;
- in response to detecting the occurrence of the trigger, call the first AI agent and provide context associated with the trigger to the first AI agent;
- and
- receive, from the first AI agent, an output resulting from execution of the first set of instructions by the first AI agent,
- wherein the first AI agent generates the output by:
- (i) determining a semantic meaning of the context associated with the trigger, (ii) comparing the semantic meaning to one or more descriptions associated with one or more tools from a set of tools, (iii) based on the comparing, determining a list of usable tools from the set of tools for modifying a database, (iv) based on the list of usable tools and the semantic meaning, generating a first plan including the first set of instructions, and (v) executing one or more instructions from the first set of instructions.
9 more elements
A method comprising:
- configuring an automation for a first artificial intelligence (AI) agent,
- wherein the automation includes a trigger that automatically initiates a call of the first AI agent, and
- wherein calling the first AI agent causes the first AI agent to execute a first set of instructions;
- detecting an occurrence of the trigger;
- in response to detecting the occurrence of the trigger, calling the first AI agent and provide context associated with the trigger to the first AI agent;
- and
- receiving, from the first AI agent, an output resulting from execution of the first set of instructions by the first AI agent,
- wherein the first AI agent generates the output by:
- (i) determining a semantic meaning of the context associated with the trigger, (ii) comparing the semantic meaning to one or more descriptions associated with one or more tools from a set of tools, (iii) based on the comparing, determining a list of usable tools from the set of tools for modifying a database, (iv) based on the list of usable tools and the semantic meaning, generating a first plan including the first set of instructions, and (v) executing one or more instructions from the first set of instructions.
6 more elements
Dependencies are inferred from the claim text.