Back to results
US 12,755,291 B2Granted
Multi-agent generative AI systems
AssigneeThe Boston Consulting Group, Inc.
InventorsAndrea Gao, Urvi Awasthi, Sanjay Elangovan, Matthew Kropp
Life of this patent
Estimating the term…
Abstract
A multi-agent generative AI system is described. The system includes at least one data storage device storing a vector database, in which the vector database includes (i) a plurality of documents, and (ii) for each document of a plurality of documents, a respective document embedding that represents a semantic meaning of the document; at least one processor; and a memory communicatively coupled to the at least one processor, the memory storing instructions which, when executed by the at least one processor, cause the at least one processor to implement a computerized chat agent, a computerized retrieval agent, and a computerized triage agent.
Claims20
A multi-agent system for automated chatting, the system comprising:
- at least one data storage device storing a vector database, wherein the vector database
- comprises (i) a plurality of documents, and (ii) for each document of a plurality of documents, a respective document embedding that represents a semantic meaning of the document;
- at least one processor;
- and
- a memory communicatively coupled to the at least one processor, the memory storing instructions which, when executed by the at least one processor, cause the at least one processor to implement a computerized generative artificial intelligence (AI) chat agent, a computerized generative AI retrieval agent, and a computerized generative AI triage agent;
- wherein the computerized generative AI triage agent is configured for communication with the computerized generative AI chat agent, the computerized generative AI retrieval agent and a large language model (LLM) including a generative transformer model with an encoder and a decoder;
- wherein the computerized generative AI triage agent is configured to guide a conversation between the computerized generative AI chat agent and a user of a user computing device displaying a chat window of the computerized generative AI chat agent by conducting reasoning on user input of the user;
- wherein the computerized generative AI chat agent is configured to:
- receive a user input at a chat window displayed at a user interface of a user computing device, and
- transmit at least the user input to the computerized generative AI triage agent;
- wherein the computerized generative AI triage agent is configured to:
- process at least the user input to create a search query, and
- transmit the search query to the computerized generative AI retrieval agent;
- wherein the computerized generative AI retrieval agent is configured to:
- convert the search query received from the computerized generative AI triage agent into a query embedding,
- query the vector database using the query embedding to retrieve one or more documents, each of the one or more documents having a respective document embedding that matches the query embedding, and
- send the retrieved one or more documents to the computerized generative AI triage agent;
- wherein the computerized generative AI triage agent is further configured to:
- based on the one or more documents, query the LLM for information to augment domain-specific knowledge contained in the one or more documents retrieved from the vector database by the computerized generative AI retrieval agent;
- based at least on content of the one or more documents retrieved by the computerized AI generative retrieval agent and the information queried from the LLM, generate one or more instructions for a logical direction of a conversation between the user and the computerized generative AI chat agent, with a logical direction specifying a next step in the conversation between the user and the computerized generative AI chat agent, and
- transmit the one or more instructions to the computerized generative AI chat agent;
- wherein the computerized generative AI chat agent is further configured to:
- generate a response to the user input based on the one or more instructions, and
- display the response in the chat window at the user interface;
- wherein each of the computerized generative AI chat agent, the computerized generative AI triage agent and the computerized generative AI retrieval agent is distinct from each other;
- wherein the combination of the computerized generative AI chat agent, the computerized generative AI triage agent and the computerized generative AI retrieval agent reduces an amount of computational resources and data storage that would otherwise be required by a generative AI system to prevent hallucination.
23 more elements
A system for automated chatting, the system comprising:
- at least one data storage device storing a vector database, wherein the vector database comprises (i) a plurality of documents, and (ii) for each document of a plurality of documents, a respective document embedding that represents a semantic meaning of the document;
- at least one processor;
- and
- a memory communicatively coupled to the at least one processor, the memory storing instructions which, when executed by the at least one processor, cause the at least one processor to implement a computerized generative artificial intelligence (AI) chat agent, a computerized generative AI retrieval agent, and a computerized generative AI triage agent;
- wherein the computerized generative AI triage agent is configured for communication with the computerized generative AI chat agent, the computerized generative AI retrieval agent and a large language model (LLM) including a generative transformer model with an encoder and a decoder;
- wherein the computerized generative AI triage agent is configured to guide a conversation between the computerized generative AI chat agent and a user of a user computing device displaying a chat window of the computerized generative AI chat agent by conducting reasoning on user input of the user;
- wherein the computerized generative AI chat agent is configured to:
- receive a user input at a chat window displayed at a user interface of a user computing device, wherein the user input relates to a technical problem associated with a product or a service, and
- transmit the user input to the computerized generative AI triage agent;
- wherein the computerized generative AI triage agent is configured to:
- determine, based on the user input, a list of possible diagnoses and root causes of the technical problem,
- transmit the list of possible diagnoses and root causes to the computerized generative AI chat agent for determining a diagnose and a root cause of the technical problem,
- receive, from the computerized generative AI chat agent, a determination of the diagnosis and the root cause of the technical problem,
- generate a technical summary of the diagnosis and root cause of the technical problem, and
- transmit the technical summary to the computerized generative AI retrieval agent;
- wherein the computerized generative AI retrieval agent is configured to:
- convert the technical summary into a query embedding,
- query the vector database using the query embedding to retrieve one or more documents, each of the one or more documents having a respective document embedding that matches the query embedding, and
- send the retrieved one or more documents to the computerized generative AI triage agent;
- wherein the computerized generative AI triage agent is further configured to:
- based on the one or more documents, query the LLM for information to augment domain-specific knowledge contained in the one or more documents retrieved from the vector database by the computerized generative AI retrieval agent;
- based at least on content of the one or more documents retrieved by the computerized AI generative retrieval agent and the information queried from the LLM, generate one or more instructions for a logical direction of a conversation between the user and the computerized generative AI chat agent, with a logical direction specifying a next step in the conversation between the user and the computerized generative AI chat agent, and
- transmit the one or more instructions to the computerized generative AI chat agent;
- wherein the computerized generative AI chat agent is further configured to:
- based at least on content of the one or more documents, generate a response to the user input based on the one or more instructions, and
- display the response in the chat window at the user interface;
- wherein each of the computerized generative AI chat agent, the computerized generative AI triage agent and the computerized generative AI retrieval agent is distinct from each other;
- wherein the combination of the computerized generative AI chat agent, the computerized generative AI triage agent and the computerized generative AI retrieval agent reduces an amount of computational resources and data storage that would otherwise be required by a generative AI system to prevent hallucination.
25 more elements
A method implemented by multi-agent system for automated chatting, comprising:
- storing, by at least one data storage device, a vector database, wherein the vector database
- comprises (i) a plurality of documents, and (ii) for each document of a plurality of documents, a respective document embedding that represents a semantic meaning of the document;
- storing instructions which, when executed by at least one processor, cause the at least one processor to implement a computerized generative artificial intelligence (AI) chat agent, a computerized generative AI retrieval agent, and a computerized generative AI triage agent;
- wherein the computerized generative AI triage agent is configured for communication with the computerized generative AI chat agent, the computerized generative AI retrieval agent and a large language model (LLM) including a generative transformer model with an encoder and a decoder;
- wherein the computerized generative AI triage agent guides a conversation between the computerized generative AI chat agent and a user of a user computing device displaying a chat window of the computerized generative AI chat agent by conducting reasoning on user input of the user;
- receiving, by the computerized generative AI chat agent, a user input at a chat window displayed at a user interface of a user computing device,
- transmitting, by the computerized generative AI chat agent, at least the user input to the computerized generative AI triage agent;
- processing, by the computerized generative AI triage agent, at least the user input to create a search query, and
- transmitting, by the computerized generative AI triage agent, the search query to the computerized generative AI retrieval agent;
- converting, by the computerized generative AI retrieval agent, the search query received from the computerized generative AI triage agent into a query embedding;
- querying, by the computerized generative AI retrieval agent, the vector database using the query embedding to retrieve one or more documents, each of the one or more documents having a respective document embedding that matches the query embedding;
- sending, by the computerized generative AI retrieval agent, the retrieved one or more documents to the computerized generative AI triage agent;
- based on the one or more documents, querying, by the computerized generative AI triage agent, the LLM for information to augment domain-specific knowledge contained in the one or more documents retrieved from the vector database by the computerized generative AI retrieval agent;
- based at least on content of the one or more documents retrieved by the computerized AI generative retrieval agent and the information queried from the LLM, generating, by the computerized generative AI triage agent, one or more instructions for a logical direction of a conversation between the user and the computerized generative AI chat agent, with a logical direction specifying a next step in the conversation between the user and the computerized generative AI chat agent;
- transmitting, by the computerized generative AI triage agent, the one or more instructions to the computerized generative AI chat agent;
- generating by the computerized generative AI chat agent, a response to the user input based on the one or more instructions, and display the response in the chat window at the user interface;
- wherein each of the computerized generative AI chat agent, the computerized generative AI triage agent and the computerized generative AI retrieval agent is distinct from each other;
- wherein the combination of the computerized generative AI chat agent, the computerized generative AI triage agent and the computerized generative AI retrieval agent reduces an amount of computational resources and data storage that would otherwise be required by a generative AI system to prevent hallucination.
15 more elements
One or more machine-readable storage devices storing instructions that are executable by one or more processors to perform operations comprising:
- storing, by at least one data storage device, a vector database, wherein the vector database
- comprises (i) a plurality of documents, and (ii) for each document of a plurality of documents, a respective document embedding that represents a semantic meaning of the document;
- storing instructions which, when executed by at least one processor, cause the at least one processor to implement a computerized generative artificial intelligence (AI) chat agent, a computerized generative AI retrieval agent, and a computerized generative AI triage agent;
- wherein the computerized generative AI triage agent is configured for communication with the computerized generative AI chat agent, the computerized generative AI retrieval agent and a large language model (LLM) including a generative transformer model with an encoder and a decoder;
- wherein the computerized generative AI triage agent guides a conversation between the computerized generative AI chat agent and a user of a user computing device displaying a chat window of the computerized generative AI chat agent by conducting reasoning on user input of the user;
- receiving, by the computerized generative AI chat agent, a user input at a chat window displayed at a user interface of a user computing device,
- transmitting, by the computerized generative AI chat agent, at least the user input to the computerized generative AI triage agent;
- processing, by the computerized generative AI triage agent, at least the user input to create a search query, and
- transmitting, by the computerized generative AI triage agent, the search query to the computerized generative AI retrieval agent;
- converting, by the computerized generative AI retrieval agent, the search query received from the computerized generative AI triage agent into a query embedding;
- querying, by the computerized generative AI retrieval agent, the vector database using the query embedding to retrieve one or more documents, each of the one or more documents having a respective document embedding that matches the query embedding;
- sending, by the computerized generative AI retrieval agent, the retrieved one or more documents to the computerized generative AI triage agent;
- based on the one or more documents, querying, by the computerized generative AI triage agent, the LLM for information to augment domain-specific knowledge contained in the one or more documents retrieved from the vector database by the computerized generative AI retrieval agent;
- based at least on content of the one or more documents retrieved by the computerized AI generative retrieval agent and the information queried from the LLM, generating, by the computerized generative AI triage agent, one or more instructions for a logical direction of a conversation between the user and the computerized generative AI chat agent, with a logical direction specifying a next step in the conversation between the user and the computerized generative AI chat agent;
- transmitting, by the computerized generative AI triage agent, the one or more instructions to the computerized generative AI chat agent;
- generating by the computerized generative AI chat agent, a response to the user input based on the one or more instructions, and display the response in the chat window at the user interface;
- wherein each of the computerized generative AI chat agent, the computerized generative AI triage agent and the computerized generative AI retrieval agent is distinct from each other;
- wherein the combination of the computerized generative AI chat agent, the computerized generative AI triage agent and the computerized generative AI retrieval agent reduces an amount of computational resources and data storage that would otherwise be required by a generative AI system to prevent hallucination.
15 more elements
Dependencies are inferred from the claim text.