US-12627625-B1 - Dynamic linking backend graph nodes for communication session representation

US12627625B1US 12627625 B1US12627625 B1US 12627625B1US-12627625-B1

Abstract

In a communication session, users start a chat, leave the conversation, and return later to continue chatting. The system enables dynamic linking of electronic conversations that may be out of chronological order. The system projects transcripts from earlier occurring conversations as a single long running communication session. The system uses graphs and immutable nodes to model a chain of conversations to recreate the long running communication session. Users resume previous conversations with the context, metadata, and transcripts carried forward. Users do not need to repeat themselves when they return to a chat and agents have access to the entire conversation history and metadata.

Inventors

Rahul Ajay Nafde
Anshul Arun Agrawal
Pawel Boguszewski
Hitesh Kathuria
Shivi Verma
Sachin Shahurao Chitale
Sharon Holmes
Charles Salmon
Thayn Moore
Joseph Sullivan

Assignees

AMAZON TECHNOLOGIES, INC.

Dates

Publication Date: 20260512
Application Date: 20230630

Claims (16)

1 . A computer-implemented method comprising: receiving a first request to create a new interaction between a user profile and an agent profile; causing presentation, in a first graphical user interface, of a current communication session between the user profile and the agent profile; causing presentation, in the first graphical user interface, of text data received from an agent computing device associated with the agent profile; receiving a second request to link the new interaction to a previous interaction, wherein the second request comprises a selection of a previous interaction among a plurality of previous interactions; in response to receiving the second request, creating a new interaction node, and adding a link to the new interaction node and a previous interaction node for the previous interaction, wherein a graph comprises the link, the new interaction node, and the previous interaction node; receiving, via the first graphical user interface, a request for a previous communication session associated with the current communication session; loading a threshold number of a plurality of previous nodes linked to the new interaction node; identifying the previous interaction node associated with the new interaction node based at least in part on the link, wherein identifying the previous interaction node further comprises: traversing the link within the plurality of previous nodes from the new interaction node to the previous interaction node; and selecting the previous interaction node from the plurality of previous nodes; receiving a transcript associated with the previous interaction node; and causing presentation, in the first graphical user interface, of the transcript, wherein causing presentation of the transcript further comprises: linking text from the transcript to text from the current communication session for presentation as a continuous session, wherein presentation of the continuous session comprises presentation, in a same graphical user interface chat window, of the text from the transcript in a continuous stream from the text from the current communication session.
2 . The computer-implemented method of claim 1 , wherein the selection of the previous interaction comprises a previous interaction identifier, wherein adding the link comprises: adding the link to the new interaction node and the previous interaction node identified by the previous interaction identifier.
3 . The computer-implemented method of claim 1 , wherein receiving the request for the previous communication session further comprises: receiving, via the first graphical user interface, user interaction input indicating a navigation change with respect to presented text in the first graphical user interface.
4 . The computer-implemented method of claim 1 further comprising: identifying a supplemental data item associated with the previous interaction; and causing presentation, in a second graphical user interface of the agent computing device, of: (i) text from the transcript combined with text from the current communication session, and (ii) the supplemental data item.
5 . The computer-implemented method of claim 4 , wherein the supplemental data item comprises (i) a document attachment or (ii) an attribute and attribute value.
6 . A system comprising: a data storage medium to store computer-executable instructions; and a computer hardware processor in communication with the data storage medium, wherein the computer hardware processor executes the computer-executable instructions to at least: receive a request to create a new interaction between a user profile and an agent profile; cause presentation, in a graphical user interface, of a current communication session between the user profile and the agent profile; receive a request to link the new interaction to a previous interaction among a group of previous interactions; receive, via the graphical user interface, a request for a previous communication session associated with the current communication session; load a first threshold number of a plurality of previous interactions associated with the new interaction; load a second threshold number of a plurality of transcripts, wherein each transcript from the plurality of transcripts is associated with an interaction from the plurality of previous interactions; select a previous transcript from the plurality of transcripts; and cause presentation, in the graphical user interface, of the previous transcript, wherein to cause presentation of the previous transcript, the computer hardware processor executes the computer-executable instructions to at least: link text from the previous transcript to text from the current communication session for presentation as a continuous session, wherein presentation of the continuous session comprises presentation, in a same graphical user interface chat window, of the text from the previous transcript in a continuous stream from the text from the current communication session.
7 . The system of claim 6 , wherein the request to link the new interaction to the previous interaction comprises a previous interaction identifier, and wherein the computer hardware processor executes additional computer-executable instructions to at least: create a new interaction node, and add a first link to (i) a previous interaction node associated with the previous interaction identifier and (ii) the new interaction node, wherein a graph comprises the first link, the previous interaction node, and the new interaction node.
8 . The system of claim 6 , wherein the computer hardware processor executes additional computer-executable instructions to at least: determine a first interaction type for a first interaction; determine that the first interaction type is not an approved previous interaction type; determine a second interaction type for the previous interaction; determine that the second interaction type is an approved previous interaction type; and send the request to link the new interaction to the previous interaction, wherein the request to link comprises an identifier for the previous interaction.
9 . The system of claim 8 , wherein the first interaction type is a survey type of interaction.
10 . The system of claim 6 , wherein the current communication session between the user profile and the agent profile is a text-based communication session, and wherein the previous transcript is a transcript of a voice-based audio recording.
11 . A system comprising: a data storage medium to store computer-executable instructions; and a computer hardware processor in communication with the data storage medium, wherein the computer hardware processor executes the computer-executable instructions to at least: receive a request to create a new interaction between a user profile and an agent profile; cause presentation, in a first graphical user interface, of a current communication session between the user profile and the agent profile; receive a request to link the new interaction to a first previous interaction among a plurality of previous interactions, wherein the request to link the new interaction to the first previous interaction comprises a previous interaction identifier; create a new interaction node; identify an initial interaction node identified by the previous interaction identifier; identify a farthest interaction node linked to the initial interaction node; add a link to the new interaction node and the farthest interaction node, wherein a graph comprises the new interaction node, the farthest interaction node, and the link; receive, via the first graphical user interface, a request for a previous communication session associated with the current communication session; identify a second previous interaction associated with the new interaction; receive a previous transcript associated with the second previous interaction; and cause presentation, in the first graphical user interface, of the previous transcript, wherein to cause presentation of the previous transcript, the computer hardware processor executes the computer-executable instructions to at least: link text from the previous transcript to text from the current communication session for presentation as a continuous session, wherein presentation as the continuous session comprises presentation, in a same graphical user interface chat window, of the text from the previous transcript in a continuous stream from the text from the current communication session.
12 . The system of claim 11 , wherein the computer hardware processor executes additional computer-executable instructions to at least: cause presentation, in the first graphical user interface, of first text data received from a user computing device associated with the user profile; and cause presentation, in the first graphical user interface, of second text data received from an agent computing device associated with the agent profile.
13 . The system of claim 11 , wherein to identify the second previous interaction, the computer hardware processor executes further computer-executable instructions to: traverse the link from the new interaction node to the farthest interaction node.
14 . The system of claim 11 , wherein the computer hardware processor executes further computer-executable instructions to at least: load a threshold number of a plurality of previous nodes linked to the new interaction node, wherein to identify the second previous interaction, the computer hardware processor executes the further computer-executable instructions to: traverse the link within the plurality of previous nodes from the new interaction node to the farthest interaction node; and select the farthest interaction node from the plurality of previous nodes.
15 . The system of claim 11 , wherein the computer hardware processor executes further computer-executable instructions to at least: load a first threshold number of a plurality of previous nodes linked to the new interaction node; and load a second threshold number of a plurality of transcripts, wherein each transcript from the plurality of transcripts is associated with a node from the plurality of previous nodes, wherein to identify the previous transcript, the computer hardware processor executes the further computer-executable instructions to: select the previous transcript from the plurality of transcripts.
16 . The system of claim 11 , wherein the computer hardware processor executes additional computer-executable instructions to at least: determine a first interaction type for a first interaction; determine that the first interaction type is not an approved previous interaction type; determine a second interaction type for the first previous interaction; and determine that the second interaction type is an approved previous interaction type, wherein the first previous interaction is the second previous interaction.

Description

BACKGROUND In a chat context, database retrieval can be used to retrieve and present messages. Typically, series of chat messages between two users are stored in a database. A chat message can have a sender, a recipient, and a timestamp. During a chat conversation between two users, a user can navigate in a user interface to view previous messages between the users. The system can retrieve messages between the two users in a reverse chronological order based on the timestamps. However, such systems lack several functions, such as, but not limited to, the ability to link previous conversations out of chronological order and the ability to link previous conversations from different users. In other chat contexts, a user can communicate with an agent. In some existing chat solutions, a user can drop in and out of the electronic conversation. When a user returns to an agent chat session there are a lot of different potential interactions with which the user could be presented. If a user is presented their most recent user-agent chat session, such a presentation may not be the most relevant to the issue at hand. BRIEF DESCRIPTION OF THE DRAWINGS These and other features, aspects, and advantages are described below with reference to the drawings, which are intended for illustrative purposes and should in no way be interpreted as limiting the scope of the embodiments. Furthermore, various features of different disclosed embodiments can be combined to form additional embodiments, which are part of this disclosure. In the drawings, like reference characters can denote corresponding features throughout similar embodiments. The following is a brief description of each of the drawings. FIG. 1 is a schematic block diagram depicting an illustrative network environment for implementing a user-agent interaction system. FIG. 2 is a schematic block diagram depicting an illustrative general architecture of a computing system for implementing the user-agent interaction system referenced in the network environment depicted in FIG. 1. FIG. 3 is a schematic diagram depicting an interaction model used in some systems. FIG. 4 is a schematic diagram depicting an improved interaction model. FIGS. 5A-5C depict illustrative communication graphical user interfaces. FIG. 6 is a flow diagram depicting illustrative interactions of a user-agent interaction system. FIG. 7 is a flow chart depicting a method implemented by the user-agent interaction system for enabling persistent communication sessions. DETAILED DESCRIPTION Generally described, aspects of the present disclosure are directed to improved linking and retrieving of electronic communications between at least two users, such as between a customer and a customer service agent. As described above, electronic message database retrieval typically involves retrieving electronic messages between user profiles in reverse chronological order and by timestamp. These systems do not allow linking of electronic messages out of chronological order and/or do not allow linking of electronic messages that can be from different agent profiles. As described herein, out of chronological order can mean that two interactions, e.g., between a user and an agent, can be directly linked where there may be intermediary interactions that occurred between the two interactions but that are omitted from the linked interactions. An improved system can allow dynamic linking of backend interaction nodes, which can be out of chronological order, that can enable a seamless presentation of transcripts between user profiles and agent profiles. The backend system can maintain multiple discrete transcripts for each interaction. As a user reviews their conversation history (such as by navigating in the graphical user interface), the backend system can dynamically identify previous interactions and associated transcripts and load those transcripts for presentation to the user such that the user perceives a continuous transcript. From an agent perspective, when a user returns, the agent can view the entire conversation history and associated metadata. The system can support loading any previous interaction and transcript, and linking a current interaction with any previous interaction. The system can receive a request to link an interaction with any previous interaction among multiple different previous interactions. The system can also support linking interactions and transcripts from different communication channels, such as linking voice and chat interactions and their corresponding transcripts into a continuous experience. With a persistent communication session, users can resume previous conversations with the context, metadata, and transcripts carried forward. Users may not need to repeat themselves when they return to a communication session and agents have access to the entire conversation history and context. The systems and methods described herein may improve computer performance. Some existing systems can load tr