# US 11,907,920 B2 · User interaction artificial intelligence chat engine for integration of automated machine generated responses > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 11,907,920 B2 - Document ID: US11907920B2 - Status: Granted - Kind: B2 - Application number: 17/661,388 - Filed: 2022-04-29 - Published: 2023-11-02 - Granted: 2024-02-20 - Earliest claimed: 2022-04-29 - Assignees: Truist Bank - Applicants: Truist Bank - Inventors: Sudeshna Banerjee; Paul Gerard Mistor - Web page: https://patentgrep.com/patent/US11907920B2 ## CPC - G06Q 20/108: {Remote banking, e.g. home banking} - H04M 3/527: Centralised call answering arrangements not requiring operator intervention - G06F 16/35: Clustering; Classification - G06F 40/20: Natural language analysis (semantic analysis of natural language G06F40/30) - G06F 40/30: Semantic analysis - G06F 40/35: Discourse or dialogue representation - G06N 3/04: Architecture, e.g. interconnection topology - G06N 3/044: Recurrent networks, e.g. Hopfield networks - G06N 3/045: Combinations of networks - G06N 3/0464: Convolutional networks [CNN, ConvNet] - G06N 3/09: Supervised learning - G06Q 30/015: Providing customer assistance, e.g. assisting a customer within a business location or via helpdesk - G06Q 30/06: Buying, selling or leasing transactions - G06Q 40/02: Banking, e.g. interest calculation or account maintenance (credit or loans G06Q40/03) - G10L 15/22: Procedures used during a speech recognition process, e.g. man-machine dialogue - H04L 51/02: using automatic reactions or user delegation, e.g. automatic replies or chatbot-generated messages - H04L 51/046: Interoperability with other network applications or services - H04M 3/493: Interactive information services, e.g. directory enquiries {; Arrangements therefor, e.g. interactive voice response [IVR] systems or voice portals} - H04M 3/5166: {in combination with interactive voice response systems or voice portals, e.g. as front-ends} - G10L 2015/223: {Execution procedure of a spoken command} - H04L 51/04: Real-time or near real-time messaging, e.g. instant messaging [IM] - H04M 3/2281: {Call monitoring, e.g. for law enforcement purposes; Call tracing; Detection or prevention of malicious calls} - H04M 3/4936: {Speech interaction details (speech recognition per se G10L15/00)} - G06Q: INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR - H04M: TELEPHONIC COMMUNICATION (circuits for controlling other apparatus via a telephone cable and not involving telephone switching apparatus G08) - G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N) - G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS - G10L: SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING - H04L: TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION (arrangements common to telegraphic and telephonic communication H04M) ## Abstract An artificial intelligence (AI) system for guiding a user interaction in a phone call or chat session. The system includes a computer running an AI algorithm, such as a machine learning algorithm, which is trained to recognize patterns in user interaction dialog which lead to satisfactory outcomes for the user. The system may operate in a completely autonomous mode, or the system connect a human agent in the loop. The algorithm adaptively guides the dialog to achieve a favorable outcome based on the current status of the dialog—including identifying a next question to ask, information to provide, or an action to take. The algorithm is trained via supervised learning using real dialog transcriptions from past user interactions which have been supplemented with decision points and outcomes. After deployment, update training may be performed on the algorithm using data captured by the system after the user interactions. ## Claims ### Claim 1 A system for guiding a user interaction during a phone call or a chat session, said system comprising: a computer with one or more processor and memory, where the computer executes a machine learning algorithm configured to guide dialog and actions during a phone call or a chat session with a user concerning a user matter; and a network connection operatively connecting the user via a user device to the computer, where the machine learning algorithm is configured to perform steps including: asking an initial question of the user and receiving a response from the user; determining a next step based on the response, where the next step is determined by the machine learning algorithm based on probability of a most rapid resolution of the user matter, and where the next step is selected from a group consisting of asking another question of the user and performing an action; determining a next question when the next step is asking another question of the user; determining a next action when the next step is performing an action, where the next action is selected from a group comprising consisting of authenticating an identity of the user, providing account information to the user, performing an account transaction for the user, providing assistance to the user in logging into an electronic data system, and connecting a live agent; connecting a live agent into the phone call or the chat session when connecting the live agent is determined as the next action, where connecting the live agent is determined as the next action for reasons including ambiguity in the response from the user and complexity of the user matter; presenting the next question or the next action to the user; receiving a user response to the next question or next action which was presented; determining if the user matter has been resolved based on the user response; returning to determining a next step when the user matter has not been resolved; and ending the phone call or the chat session and storing data describing the user interaction when the user matter has been resolved. ### Claim 2 (depends on 1) The system according to claim 1 wherein the phone call or the chat session transpires between the user device and the computer over the network connection, where the user device is one of a mobile phone, a non-mobile phone, a tablet device, a computer or a display screen with a virtual or physical keyboard. ### Claim 3 (depends on 2) The system according to claim 2 wherein the network connection includes one or more of the Internet, a local area network, a wide area network, a cellular network, a landline or hard-wired telephone system, and a satellite communication system. ### Claim 4 (depends on 1) The system according to claim 1 wherein the machine learning algorithm uses a neural network. ### Claim 5 (depends on 4) The system according to claim 4 wherein the neural network is a feedforward neural network, a convolutional neural network or a recurrent neural network. ### Claim 6 (depends on 4) The system according to claim 4 wherein at least part of the neural network is configured as a natural language processing application. ### Claim 7 (depends on 1) The system according to claim 1 wherein the machine learning algorithm is initially trained via supervised learning using a dataset of previous user interactions supplemented with decision points, actions which resulted in matter resolution, and user satisfaction level. ### Claim 8 (depends on 7) The system according to claim 7 wherein the machine learning algorithm is periodically provided with update training via supervised learning using a dataset containing the data describing each user interaction which was stored when each phone call or chat session was ended. ### Claim 9 (depends on 1) The system according to claim 1 wherein determining a next step includes considering, by the machine learning algorithm, one or more previous user responses from the user interaction in addition to a most recent user response. ### Claim 10 (depends on 1) The system according to claim 1 wherein presenting the next question or the next action to the user includes presenting by the computer when no live agent is participating in the phone call or chat session, and presenting by the live agent when the live agent is participating. ### Claim 11 A system for guiding a user interaction during a phone call or a chat session, said system comprising: a computer with one or more processor and memory, where the computer executes a machine learning algorithm configured to guide dialog and actions during a phone call or a chat session with a user concerning a user matter; and a network connection operatively connecting the user via a user device to the computer, where the machine learning algorithm includes a neural network configured to perform steps including: asking an initial question of the user and receiving a response from the user; determining a next step based on the response, where the next step is determined by the machine learning algorithm based on probability of a most rapid resolution of the user matter, and where the next step is selected from a group consisting of asking another question of the user and performing an action; determining a next question when the next step is asking another question of the user; determining a next action when the next step is performing an action, where the next action is selected from a group consisting of authenticating an identity of the user, providing account information to the user, performing an account transaction for the user, providing assistance to the user in logging into an electronic data system, and connecting a live agent; connecting a live agent into the phone call or the chat session when connecting the live agent is determined as the next action, where connecting the live agent is determined as the next action for reasons including ambiguity in the response from the user and complexity of the user matter; presenting the next question or the next action to the user; receiving a user response to the next question or next action which was presented; determining if the user matter has been resolved based on the user response; returning to determining a next step when the user matter has not been resolved; and ending the phone call or the chat session and storing data describing the user interaction when the user matter has been resolved, and where the neural network is initially trained via supervised learning using a dataset of previous user interactions supplemented with decision points, actions which resulted in matter resolution, and user satisfaction level. ### Claim 12 (depends on 11) The system according to claim 11 wherein the neural network is periodically provided with update training via supervised learning using a dataset containing the data describing each user interaction which was stored when each phone call or chat session was ended. ### Claim 13 (depends on 11) The system according to claim 11 wherein determining a next step includes considering, by the neural network, one or more previous user responses from the user interaction in addition to a most recent user response. ### Claim 14 A method for guiding a user interaction during a phone call or a chat session, said method comprising: providing a computer with one or more processor and memory, where the computer executes a machine learning algorithm configured to guide dialog and actions during a phone call or a chat session with a user concerning a user matter, and the computer uses a network connection to connect to the user via a user device, where the machine learning algorithm is configured to perform steps including: asking an initial question of the user and receiving a response from the user; determining a next step based on the response, where the next step is determined by the machine learning algorithm based on probability of a most rapid resolution of the user matter, and where the next step is selected from a group consisting of asking another question of the user and performing an action; determining a next question when the next step is asking another question of the user; determining a next action when the next step is performing an action, where the next action is selected from a group comprising consisting of authenticating an identity of the user, providing account information to the user, performing an account transaction for the user, providing assistance to the user in logging into an electronic data system, and connecting a live agent; connecting a live agent into the phone call or the chat session when connecting the live agent is determined as the next action, where connecting the live agent is determined as the next action for reasons including ambiguity in the response from the user and complexity of the user matter; presenting the next question or the next action to the user; receiving a user response to the next question or next action which was presented; determining if the user matter has been resolved based on the user response; returning to determining a next step when the user matter has not been resolved; and ending the phone call or the chat session and storing data describing the user interaction when the user matter has been resolved. ### Claim 15 (depends on 14) The method according to claim 14 wherein the machine learning algorithm uses a neural network. ### Claim 16 (depends on 15) The method according to claim 15 wherein the neural network is a feedforward neural network, a convolutional neural network or a recurrent neural network, and at least part of the neural network is configured as a natural language processing application. ### Claim 17 (depends on 14) The method according to claim 14 wherein the machine learning algorithm is initially trained via supervised learning using a dataset of previous user interactions supplemented with decision points, actions which resulted in matter resolution, and user satisfaction level. ### Claim 18 (depends on 14) The method according to claim 14 wherein determining a next step includes considering, by the machine learning algorithm, one or more previous user responses from the user interaction in addition to a most recent user response. ### Claim 19 (depends on 14) The method according to claim 14 wherein presenting the next question or the next action to the user includes presenting by the computer when no live agent is participating in the phone call or chat session, and presenting by the live agent when the live agent is participating.