Methods and systems for generating automated replies for conversational interactions

Abstract

Methods and server systems for generating automated replies for conversational interactions from clients are described herein Method performed by server system includes receiving, from a first Machine Learning (ML) model, client intent information for a conversational interaction and performing, by second ML model(s), a set of operations. As a part of the set of operations, method includes detecting presence of sub-intent(s) in the client intent information, upon detecting one sub-intent, assigning a specific automatable process to the conversational interaction. Method includes extracting multiple relevant entities from the conversational interaction based on relevant text(s) from the conversational interaction. Method also includes accessing a list of tasks associated with the specific automatable process and generating query resolution information based on processing the client intent information for the list of tasks. Method also includes generating conversational interaction response template(s) for the conversational interaction based on the multiple relevant entities, and query resolution information.

Description

METHODS AND SYSTEMS FOR GENERATING AUTOMATED REPLIES FOR CONVERSATIONAL INTERACTIONSTECHNICAL FIELD

[0001] The present disclosure relates to the field of customer service in a conversation ecosystem and, more particularly, to electronic methods and complex processing systems for generating automated replies for conversational interactions with users or clients.BACKGROUND

[0002] Customer service is vital for all businesses since it can act as the differentiating factor between a successful business and an unsuccessful one. To ensure good customer service, businesses employ teams of agents who may be tasked with proactively reaching out to customers to solve their problems. In addition, when customers reach out to the customer service team with their problems, the agents are responsible for addressing these problems. Nowadays, it has become essential for businesses to provide good customer service by quickly and effectively resolving customer issues to ensure repeat customers and good word-of-mouth marketing. Generally, businesses ensure that their customers can raise their queries or concerns through various interaction channels such as phone calls, electronic mail (email), chat, social media, and the like. The interaction between the customer and the agent can be referred to as a conversational interaction. Upon receiving the query, the agent is to understand the issue faced by the customer and find a suitable solution for the said issue. In various examples, the queries can be product usage-related queries, billing -related queries, shipping-related queries, product disposal-related queries, service-related queries, and so on. It is to be noted that the process of preparing an appropriate response to resolve the customer’s query is a labor- intensive and time-consuming process due to its inherent complexity.

[0001] With the growth in the consumer base of a business, management of customer queries has become quite complex since agents have to engage in long conversations with clients over extended durations. Due to this, businesses have to either hire more agents or burden the existing staff with longer working hours to provide a good customer experience for their clients. Since it is difficult for businesses to hire and train new staff, many clients face delayed or insufficient resolutions to their queries, leading to poor customer experience. Further, as resolution time increases, the burden on the customer query management systems operated by the businesses also increases in terms of higher processing and memoryrequirements.

[0002] Thus, it is desirable to find technological solutions that improve the response time of an agent while solving a customer query.SUMMARY

[0003] There exists a need for techniques to overcome one or more limitations stated above such as the long query resolution time, poor customer experience, burden on agents of understanding and resolving the client queries, and increased load on the processing and memory requirements of the customer query management systems.

[0004] Various embodiments of the present disclosure provide methods and systems that reduce the burden on agents by providing recommendations to the agents through resolution templates. Since the burden on agents is reduced, their response time improves significantly which in turn improves the customer experience. Further, since the time spent by agents on the relevant portals of the customer query management systems is also reduced due to quick query resolution, the processing and memory requirements of the customer query management systems are also reduced, which saves on electricity costs as well. Further, as the time spent on each query resolution is reduced, now the same agent can serve more clients thereby reducing labor requirements for their employers as well.

[0005] To achieve the above and other objectives of the present disclosure, in one aspect, a computer-implemented method for generating automated replies to one or more conversational interactions from one or more clients is disclosed. The computer-implemented method is performed by a server system. The computer-implemented method includes receiving, from a first Machine Learning (ML) model, client intent information for a conversational interaction initiated by a client. The computer-implemented method further includes performing, by one or more second ML models, a set of operations. Thus, as a part of the set of operations, the method includes detecting the presence of one or more sub-intents in the client intent information. Upon detecting one sub-intent, the computer-implemented method includes assigning a specific automatable process to the conversational interaction based, at least in part, on comparing the client intent information with a set of pre-defined automatable processes. Further, the computer- implemented method includes extracting a plurality of relevant entities from the conversational interaction based, at least in part, on one or more relevant texts associated with the conversational interaction. Furthermore, the computer-implemented method includes accessing a list of tasks associated with the specificautomatable process from a database. The computer-implemented method also includes generating query resolution information for the conversational interaction based, at least in part, on processing the client intent information for the list of tasks. Finally, the computer- implemented method includes generating one or more conversational interaction response templates for the conversational interaction based, at least in part, on the plurality of relevant entities, and the query resolution information.

[0006] An advantage of some embodiments is that by receiving the client intent information for the conversational interaction from the first ML model, the complex process of understanding the context and requirement of the client query is minimized. Since the time taken by the first ML model to understand the client's intent is nearly instantaneous, the time taken to address the query from the customer is reduced.

[0007] Moreover, detecting the presence of one or more sub-intents in the client intent information helps in deciding whether a specific automatable process can be assigned to the conversational interaction or not. As may be appreciated, this aspect also helps in reducing the query resolution time. Further, accessing the set of tasks associated with the specific automatable process facilitates the generation of the query resolution information by performing the corresponding tasks. Herein, the query resolution information includes information related to an appropriate resolution to a problem faced by the client as mentioned in the conversational interaction.

[0008] In an aspect, the step of receiving the client intent information includes receiving the conversational interaction initiated by the client. The step further includes extracting a set of relevant sections from the conversational interaction. Further, it includes generating a first relevant text of the one or more relevant texts based, at least in part, on applying a first set of preprocessing operations on the set of relevant sections. The step further includes generating, by the first ML model, the client intent information for the conversational interaction based, at least in part, on the first relevant text and a historical conversational interaction dataset.

[0009] An advantage of some embodiments is that extracting the set of relevant sections from the conversational interaction facilitates the server system to consider only relevant text from the query for processing. This aspect reduces the processing load on the server system while improving the processing speed. In addition, as unnecessary text is discarded, the resolution recommendation generated for the agent from the set of relevantsections is accurate and precise.

[0010] In an aspect, the computer-implemented method further includes fine-tuning, by the first ML model, the client intent information based, at least in part, on the first relevant text and a first predefined ruleset.

[0011] An advantage of some embodiments is that fine-tuning the client intent information helps the server system in scrutinizing the actual intent of the client thereby leading to improved results.

[0012] In an aspect, the step of extracting the plurality of relevant entities includes extracting a first set of relevant entities of the plurality of relevant entities. For extracting the first set of relevant entities, it is required for the computer-implemented method to generate a second relevant text of the one or more relevant texts based, at least in part, on applying a second set of preprocessing operations on the first relevant text. The computer-implemented method further includes processing, by the one or more second ML models, the second relevant text to generate the first set of relevant entities based, at least in part, on the second relevant text and a second predefined ruleset.

[0013] An advantage of some embodiments is that by preprocessing the first relevant text, the second relevant text required for efficiently extracting the first set of relevant entities can be obtained.

[0014] In an aspect, the step of extracting the plurality of relevant entities includes extracting a second set of relevant entities of the plurality of relevant entities. For extracting the second set of relevant entities, it is required for the computer-implemented method to generate a third relevant text of the one or more relevant texts based, at least in part, on applying a third set of preprocessing operations on the first relevant text. The computer-implemented method further includes processing, via the one or more second ML models, the first relevant text to generate an intermediate set of entities based, at least in part, on a third predefined ruleset. Further, the computer-implemented method includes comparing the third relevant text with the intermediate set of entities to generate the second set of relevant entities.

[0015] An advantage of some embodiments is that by preprocessing the first relevant text, the third relevant text required for efficiently extracting the second set of relevant entities can be obtained.

[0016] In an aspect, the step of processing the client intent information for the list of tasks includes performing each task from the list of tasks to generate the query resolutioninformation for the conversational interaction.

[0017] An advantage of some embodiments is that performing the list of tasks via the server system eliminates the requirement of any kind of manual intervention from the agent. This aspect further helps in reducing the query resolution time.

[0018] In an aspect, the step of detecting the presence of the one or more sub-intents in the client intent information includes detecting two or more sub-intents in the client intent information based, at least in part, on an identification of one or more conjunction texts in the client intent information. In an aspect, the computer-implemented method further includes assigning a non- automatable process to the conversational interaction based at least on the client intent information being associated with at least two or more sub-intents.

[0019] An advantage of some embodiments is that by assigning the non-automatable process to the conversational interaction, the conversational interaction is discarded from automatability consideration. This enables the agent to take over the query resolution process for manual processing without wasting the processing resources of the server system. This aspect acts as a fail-safe step while ensuring that only correct information is shared with the client, thus preventing irrelevant information from being pushed to the client.

[0020] As per another embodiment of the present disclosure, a server system is disclosed. The server system includes a communication interface and a memory including executable instructions. The server system also includes a processor communicably coupled to the memory. The processor is configured to execute the instructions to cause the server system, at least in part, to receive, from a first Machine Learning (ML) model, client intent information for a conversational interaction initiated by a client. The server system is further caused to perform, by one or more second ML models, a set of operations. As a part of performing the set of operations, the server system is caused to detect the presence of one or more sub-intents in the client intent information. Upon detecting one sub-intent, the server system is further caused to assign a specific automatable process to the conversational interaction based, at least in part, on comparing the client intent information with a set of pre-defined automatable processes. The server system is also caused to extract a plurality of relevant entities from the conversational interaction based, at least in part, on one or more relevant texts associated with the conversational interaction. Furthermore, the server system is caused to access a list of tasks associated with the specific automatable process from a database. Moreover, the server system is caused to generate query resolution information for the conversational interaction based, atleast in part, on processing the client intent information for the list of tasks. Finally, the server system is caused to generate one or more conversational interaction response templates for the conversational interaction based, at least in part, on the plurality of relevant entities, and the query resolution information.

[0021] As per yet another embodiment of the present disclosure, a non-transitory computer-readable storage medium is disclosed. The non-transitory computer-readable storage medium includes computer-executable instructions that, when executed by at least a processor of a server system, cause the server system to perform a method. The method includes receiving, from a first Machine Learning (ML) model, client intent information for a conversational interaction initiated by a client. The computer-implemented method further includes performing, by one or more second ML models, a set of operations. Thus, as a part of the set of operations, the method includes detecting the presence of one or more sub-intents in the client intent information. Upon detecting one sub-intent, the computer-implemented method includes assigning a specific automatable process to the conversational interaction based, at least in part, on comparing the client intent information with a set of pre-defined automatable processes. Further, the computer- implemented method includes extracting a plurality of relevant entities from the conversational interaction based, at least in part, on one or more relevant texts associated with the conversational interaction. Furthermore, the computer-implemented method includes accessing a list of tasks associated with the specific automatable process from a database. The computer-implemented method also includes generating query resolution information for the conversational interaction based, at least in part, on processing the client intent information for the list of tasks. Finally, the computer- implemented method includes generating one or more conversational interaction response templates for the conversational interaction based, at least in part, on the plurality of relevant entities, and the query resolution information.

[0022] The foregoing summary is illustrative only and is not intended to be in any way limiting. In addition to the illustrative aspects, embodiments, and features described above, further aspects, embodiments, and features will become apparent by reference to the drawings and the following detailed description.BRIEF DESCRIPTION OF FIGURES

[0023] For a more complete understanding of example embodiments of the present technology, reference is now made to the following descriptions taken in connection with the accompanying drawings in which:

[0024] FIG. 1 is an example representation of a conversational interaction environment, in accordance with various embodiments of the present disclosure;

[0025] FIG. 2 illustrates a simplified block diagram of a server system, in accordance with an embodiment of the present disclosure;

[0026] FIG. 3 is a flowchart depicting a process of generating Artificial Intelligence (Al) resolution templates for an agent in response to a client’s query email from a client, in accordance with an embodiment of the present disclosure;

[0027] FIGS. 4A and 4B, collectively illustrate a process of generating Al resolution templates for an agent in response to a specific query email from a client, in accordance with an embodiment of the present disclosure; and

[0028] FIG. 5 illustrates a flow diagram of a method of generating one or more automated replies to one or more conversational interactions from one or more clients, in accordance with an embodiment of the present disclosure.

[0029] The drawings referred to in this description are not to be understood as being drawn to scale except if specifically noted, and such drawings are only exemplary in nature.DETAILED DESCRIPTION

[0030] In the following description, for purposes of explanation, numerous specific details are set forth in order to provide a thorough understanding of the present disclosure. It will be apparent, however, to one skilled in the art that the present disclosure can be practiced without these specific details. Descriptions of well-known components and processing techniques are omitted to not obscure the embodiments herein unnecessarily. The examples used herein are intended merely to facilitate an understanding of ways in which the embodiments herein may be practiced and to further enable those of skill in the art to practice the embodiments herein. Accordingly, the examples should not be construed as limiting the scope of the embodiments herein.

[0031] References in this specification to “one embodiment” or “an embodiment” means that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least one embodiment of the present disclosure. The appearances of the phrase “in an embodiment” in various places in the specification are not necessarily all referring to the same embodiment, nor are separate or alternative embodiments mutually exclusive of other embodiments. Moreover, various features are described which may be exhibited by some embodiments and not by others. Similarly, various requirements aredescribed which may be requirements for some embodiments but not for other embodiments.

[0032] Moreover, although the following description contains many specifics for the purposes of illustration, anyone skilled in the art will appreciate that many variations and / or alterations to said details are within the scope of the present disclosure. Similarly, although many of the features of the present disclosure are described in terms of each other, or in conjunction with each other, one skilled in the art will appreciate that many of these features can be provided independently of other features. Accordingly, this description of the present disclosure is set forth without any loss of generality to, and without imposing limitations upon, the present disclosure.

[0033] Conditional language such as, among others, “can”, “could”, “might”, or “may”, unless specifically stated otherwise, are otherwise understood within the context as used in general to convey that certain embodiments include, while other embodiments do not include, certain features, elements and / or steps. Thus, such conditional language is not generally intended to imply that features, elements, and / or steps are in any way required for one or more embodiments or that one or more embodiments necessarily include logic for deciding, with or without user input or prompting, whether these features, elements and / or steps are included or are to be performed in any particular embodiment.

[0034] Disjunctive language such as the phrase “at least one of X, Y, or Z” unless specifically stated otherwise, is otherwise understood with the context as used in general to present that an item, term, etc., may be either X, Y, or Z, or any combination thereof (e.g., X, Y, and / or Z). Thus, such disjunctive language is not generally intended to, and should not, imply that certain embodiments require at least one of X, at least one of Y, or at least one of Z to each be present.

[0035] Unless otherwise explicitly stated, articles such as “a” or “an” should generally be interpreted to include one or more described items. Accordingly, phrases such as “a server system configured to” are intended to include one or more recited server systems / processors. Such one or more recited devices can also be collectively configured to carry out the stated recitations. For example, “a processor configured to carry out recitations A, B, and C” can include a first processor configured to carry out recitation A working in conjunction with a second processor configured to carry out recitations B and C. The same holds true for the use of definite articles used to introduce embodiment recitations. In addition, even if a specific number of an introduced embodiment recitation is explicitly recited, those skilled in the artwill recognize that such recitation should typically be interpreted to mean at least the recited number (e.g., the bare recitation of “two recitations” without other modifiers, typically means at least two recitations or two or more recitations).

[0036] It will be understood by those within the art that, in general, terms used herein, are generally intended as “open” terms (e.g., the term “including” or “comprising” should be interpreted as “including / comprising but not limited to,” the term “having” should be interpreted as “having at least,” the term “includes” or “comprises” should be interpreted as “includes / comprises but is not limited to,” etc.).

[0037] For expository purposes, the term ‘customer’, ‘user’, ‘client’, ‘consumer’, or ‘buyer’ (used interchangeably herein) refers to an existing or a potential user of products and / or services offered by a business or a service provider. Moreover, the term ‘customer representatives’ or ‘agents’ as used herein refers to human agents, chat-bots, self-assist systems, and, in general to any human or machine interface interacting with customers of businesses.

[0038] It is understood that conversational interactions between the agents and the clients correspond to written digital communication such as email. It plays a vital role in bridging the gap between the clients and the agents. It is expected that the agents promptly respond to the queries of the clients with resolutions that are correct and match the requirements of the clients. However, it becomes tedious for the agents as the number of clients and their queries increase. The complexity of obtaining the context of the conversation from the clients is more in scenarios where it is required for the agents to perform certain tasks. The tasks may include visiting different internal web portals to check for the status of certain processes, documents, or the like, analyzing prior trails of communication with the client, analyzing prior communication between the client and another agent in the past, and the like. Further, the agent is required to take necessary actions and reply with the most relevant information to the client to resolve the issue faced by the client.

[0039] Conventionally, various techniques are used for providing auto-resolution to client queries. In such conventional techniques, upon receiving a query (i.e., conversational interaction) from a client, a ticket is created corresponding to the received query. Later, the ticket is resolved by the agents of existing businesses based on a predefined set of resolution templates.

[0040] However, since the predefined set of resolution templates is not an exhaustivelist having resolutions for all possible client queries, sometimes the response received by the client may not be remotely close to the actual expectations of the clients. Considering that, it may be understood that existing techniques fail to understand the context and / or intent of the conversational interaction received from the clients, thereby failing to recommend an appropriate reply to the agent.

[0041] As a result, clients suffer from a bad experience since they might need to reopen their tickets again and again until they can reach the correct resolution. On the other hand, due to this bad experience, businesses end up hiring a great number of agents to manually read and resolve client queries at great expense in hiring, staffing, and equipment costs. Further, the manual nature of the process leads to slow query resolution which again negatively impacts the client experience.

[0042] To that end, various embodiments of the present disclosure aim to solve the above-mentioned technical problems by providing an approach for generating automated resolution recommendations to the agent involved in a conversational interaction with a client. The approach of the present disclosure aims to simplify the task of understanding the intent of the conversational interactions between the clients for agents. This is achieved by determining processes in the query resolution journey that are automatable in nature and automating such processes. Thus, it may be understood that the approach includes determining which of the processes of determining a resolution are automatable. Then, for conversational interactions associated with those automatable processes resolution recommendations may be generated. Further, the agent may choose one of the recommendations and respond to the client with the same resolution by making minor changes to it. This helps to optimize response time for the agents interacting with the clients.

[0043] FIG. 1 is an example representation of a conversational interaction environment 100, in accordance with various embodiments of the present disclosure. The environment 100 includes a server system 102, a plurality of agents 104A, 104B, and 104C (hereafter collectively referred to as ‘agents 104’), a plurality of clients 106A, 106B, and 106C (hereafter collectively referred to as ‘clients 106’), each coupled to, and in communication with (and / or with access to) a network 108. In a non-limiting example, the server system 102 may be configured to perform one or more operations, such as, but not limited to, generating client intent information, detecting the presence of multiple sub-intents in the client intent information, generating one or more conversational interaction response templates as a recommendation for the agents 104 to use them as a response for the received conversationalinteraction, and the like.

[0044] In an embodiment, the agents 104 may be, but are not limited to, human agents, chatbots, self-assist systems such as either web or mobile digital self-service systems, or the like that are responsible for responding to conversational interactions from the clients 106. In another embodiment, the clients 106 may be, but are not limited to, human users, representatives of an organization / business, chatbots, or the like.

[0045] The network 108 may include, without limitation, a Light Fidelity (Li-Fi) network, a Local Area Network (LAN), a Wide Area Network (WAN), a Metropolitan Area Network (MAN), a satellite network, the Internet, a fiber optic network, a coaxial cable network, an Infrared (IR) network, a Radio Frequency (RF) network, a virtual network, and / or another suitable public and / or private network capable of supporting communication among two or more of the parts or components illustrated in FIG. 1, or any combination thereof.

[0046] Various components in the environment 100 may connect to the network 108 in accordance with various wired and wireless communication protocols, such as Transmission Control Protocol / Internet Protocol (TCP / IP), User Datagram Protocol (UDP), 2ndGeneration (2G), 3rdGeneration (3G), 4thGeneration (4G), 5thGeneration (5G) communication protocols, Long Term Evolution (LTE) communication protocols, future communication protocols or any combination thereof. For example, the network 108 may include multiple different networks, such as a private network made accessible by the server system 102 and a public network (e.g., the Internet, etc.) through which the server system 102, any of the agents 104, and any of the clients 106 may communicate.

[0047] Each agent of the agents 104 is associated with an electronic device for interacting with existing / potential clients present at different geographical locations, such as clients 106. The agents 104 are depicted to be associated with different electronic devices such as desktop computers and mobile devices in the environment 100. However, it should be understood that the agents 104 may use any electronic device that may be configured to facilitate interaction between two remote individuals (for example, individuals, such as the agent 104 A and the client 106A). For example, the agent 104B may be equipped with a mobile device, such as a smartphone configured to facilitate engagement between the agent 104B and a client 106B.

[0048] Similarly, each of the clients 106 may be associated with one or more electronic devices such as, but not limited to, a mobile device (e.g., a mobile phone, a smartphone, atablet, etc.), a smart television, a laptop, a desktop for interacting with the agents 104. In the illustrated example, the client 106A is depicted to be associated with a desktop computer, the client 106B is depicted to be associated with a desktop computer as well, and the client 106C is depicted to be associated with a mobile device. Similarly, the agent 104A is depicted to be associated with a mobile device, the agent 104B is associated with a desktop computer, and the agent 104C is also associated with a desktop computer.

[0049] It should be noted that one or more clients from among the clients 106 may be associated with multiple electronic devices and may interact with the agents 104 over multiple devices and / or using multiple applications (such as Google Mail®, Outlook Mail®, Apple Business Chat®, Google Business Messaging®, Facebook Messaging®, and the like) supporting different text-based interaction channels (such as email, chat and the like). Further, multiple device-based interactions between multiple clients and multiple agents may also be executed concurrently or separately as per the client’s need.

[0050] It should be understood that the clients 106 and the agents 104 may engage in communication with each other via different applications such as, but not limited to, a mobile application, an email application, a web-based email client application, a web-based chat application, and so on, for a variety of purposes.

[0051] In an example scenario, the client 106A may initiate a conversational interaction with the agent 104A using a mobile application. The client 106A may initiate the interaction to inquire about an invoice, a bill of lading (B / L or BL), a sea waybill, or the like of a product or service offered by an employer of the agent 104A. In another example scenario, the agent 104B may initiate a conversational interaction with the client 106B related to a certain product or service as a response to an earlier client query. In such a scenario, the conversational interaction may be related to offering the client 106B to purchase the product or service, requesting feedback about the product that may be purchased by the client 106B, a resolution to a client’s query, and the like.

[0052] As described earlier, upon initiation of a conversational interaction with the client 106A, the agent 104A tries to understand the problems of the client 106A or the requirements of the client 106A. Then, the agent 104A may have to perform a set of tasks to determine a resolution for the problem of the client 106A. For example, if the client’s query is requesting the B / L, then the agent 104 A has to comprehend that the client 106 A is requesting the B / L document for a particular order made by the client 106 A, then the agent 104 A has toextract client parameters such as a customer code, a shipment Identifier (ID), or the like either from the conversational interaction or different portals associated with the main shipment portal based on client’s details such as email address or client ID. Then, the agent 104A may visit a specific portal such as a Global Customer Service System (GCSS) portal, a finance- related portal, or the like to find or generate the B / L document for the client. In one instance, the B / L document may have been previously generated and stored at the time of purchase by the client 106A and the agent 104A can now extract this from the respective platform using the client parameters. Returning to the previous example, upon finding the B / L document, the agent 104 A may verify it and share the same over email with the client 106A. As may be understood, the process undertaken by the agent 104A to fulfill the client’s requirement is tedious, time-consuming, and resource-intensive. Further, since such requirements are quite common, the agent 104A may have to perform the same process various times throughout the day, leading to inefficient use of the agent’s time. Furthermore, as many agents 104 may have to log on to such a portal and spend time on it throughout the day, the portal itself gets burdened leading to higher processing and memory requirements, which in turn leads to higher infrastructure and electricity costs.

[0053] To overcome this problem, an approach for generating resolution recommendations to the agent 104 A from an automatable process in response to a conversational interaction from the client 106A. To that end, the present disclosure describes the server system 102 that performs a plurality of operations for facilitating the generation of the resolution recommendations to the agents 104.

[0054] In one embodiment, the environment 100 may further include a database 110 coupled with the server system 102. In an example, the server system 102 coupled with the database 110 is embodied within a central server (not shown) associated with the employer of the agents 104, however, in other examples, the server system 102 can be a standalone component (acting as a hub) connected to the central server. The database 110 may be incorporated in the server system 102 or maybe an individual component connected to the server system 102 or maybe a database stored in cloud storage. In one embodiment, the database 110 may store a historical conversational interaction dataset, one or more Machine Learning (ML) models, and other necessary machine instructions required for implementing the various functionalities of the server system 102 such as firmware data, operating system, and the like. In an embodiment, the one or more ML models may include a first ML model and one or more second ML models. It is noted that the historical conversational interaction dataset,the first ML model, and the one or more second ML models have been explained in detail later in the present disclosure. In addition, the database 110 provides a storage location for data and / or metadata obtained from various operations performed by the server system 102.

[0055] In an embodiment, when a conversational interaction is initiated by a client such as the client 106A with an agent such as the agent 104A, the server system 102 is configured to receive the conversational interaction as well. In some examples, the conversational interaction may be an email, a chat message, or the like. In an instance, the conversational interaction may include a request from the client 106 A to solve a query of the client 106A. In response to this conversational interaction, the server system 102 is configured to process one or more relevant texts from the conversational interaction using the first ML model for identifying one or more client intents associated with the conversational interaction. This step results in client intent information which is further processed using the one or more second ML models. This step results in categorizing the conversational interaction in at least one process category. Herein, the process category is associated with a set of tasks that are supposed to be performed to determine a resolution for the query of the client 106A in the received conversational interaction.

[0056] In one embodiment, the server system 102 may also be configured to identify whether the process category assigned to the conversational interaction is automatable or not. Herein, if the process category is automatable, then the server system 102 can automatically implement the set of tasks that may be associated with the corresponding process category. Alternatively, the conversational interaction may be discarded from further processing if it requires the manual intervention of the agent 104A and is not completely automatable.

[0057] In some instances, the conversational interaction may be associated with multiple sub-intents. For example, the client 106 A can request the agent 104 A to check whether an invoice is generated for a particular shipment and also request the agent 104 A to provide a billing of lading. Herein, multiple sub-intents correspond to requesting for both the status of the invoice and the billing of lading. Thus, the server system 102 may be configured to detect the presence of one or more sub-intents in the client intent information using the one or more second ML models. In one scenario, upon detecting one sub-intent of the one or more sub-intents, the server system 102 may be configured to assign a specific automatable process to the conversational interaction based on comparison with a set of pre-defined automatable processes. Later, the server system 102 may extract a plurality of relevant entities from the conversational interaction based on one or more relevant texts associated with theconversational interaction. In some embodiments, some entities in the plurality of entities may correspond to entities that indicate some significant identifiers, names such as the sender’s name, and the like.

[0058] In some other embodiments, the server system 102 may be configured to access a list of tasks associated with the specific automatable process from the database 110. Further, the server system 102 may be configured to generate query resolution information for the conversational interaction based, at least in part, on processing the client intent information for the list of tasks. Herein, in an embodiment, the list of tasks may correspond to tasks that are supposed to be performed for determining a query resolution for the query in the received conversational interaction. Thus, processing the client intent information for the list of tasks may include facilitating the implementation of one or more tasks from the list of tasks. In another embodiment, the server system 102 may be configured to generate one or more conversational interaction response templates for the conversational interaction based, at least in part, on the plurality of relevant entities, and the query resolution information.

[0059] In a specific embodiment, the conversational interaction response templates may be provided to the agent 104 A as a resolution recommendation, so that the agent 104 A can choose the most appropriate template and send a reply to the client 106A in response to the conversational interaction received from the client 106A. For example, in response to receiving a query email, template emails describing the resolution for the query in the received query email are provided to the agent 104 A as a recommendation. A detailed explanation of various operations required for generating the resolution recommendations by the server system 102 is provided later in reference to FIG. 2.

[0060] Although in FIG. 1, the server system 102 is shown to be incorporated within the environment 100, in some embodiments, the server system 102 may be external to and in communication with the environment 100, for example, via the network 108. In some examples, the server system 102 may be implemented in third-party external servers to perform the various operations described herein.

[0061] The number and arrangement of systems, devices, and / or networks shown in FIG. 1 are provided as an example. There may be additional systems, devices, and / or networks; fewer systems, devices, and / or networks; different systems, devices, and / or networks; and / or differently arranged systems, devices, and / or networks than those shown in FIG. 1. Furthermore, two or more systems or devices shown in FIG. 1 may be implemented within asingle system or device, or a single system or device is shown in FIG. 1 may be implemented as multiple, distributed systems or devices. In addition, the server system 102 should be understood to be embodied in at least one computing device in communication with the network 108, which may be specifically configured, via executable instructions, to perform steps as described herein, and / or embodied in at least one non-transitory computer-readable media.

[0062] FIG. 2 illustrates a simplified block diagram of a server system 200, in accordance with an embodiment of the present disclosure. It is noted that the server system 200 may be similar to the server system 102 of FIG. 1. In one embodiment, the server system 200 is a part of the internal server operated by an organization employing the user such as the agent 104(1). In some embodiments, the server system 200 is embodied as a cloud-based and / or Software as a Service (SaaS) based architecture.

[0063] The server system 200 includes a computer system 202 and a database 204. It is noted that the database 204 is identical to the database 110 of FIG. 1. The computer system 202 includes at least one processor 206 (herein, referred to interchangeably as ‘processor 206’) for executing instructions, a memory 208, a communication interface 210, a user interface 212, and a storage interface 214 that communicates with each other via a bus 216.

[0064] In some embodiments, the database 204 is integrated into the computer system 202. For example, the computer system 202 may include one or more hard disk drives like the database 204. The user interface 212 is an interface, such as a Human Machine Interface (HMI) or a software application that allows users such as an administrator to interact with and control the server system 200 or one or more parameters associated with the server system 200. It may be noted that the user interface 212 may be composed of several components that vary based on the complexity and purpose of the application. Examples of components of the user interface 212 may include visual elements, controls, navigation, accessibility features, etc.

[0065] The storage interface 214 is any component capable of providing the processor 206 with access to the database 204. The storage interface 214 may include, for example, an Advanced Technology Attachment (ATA) adapter, a Serial ATA (SATA) adapter, a Small Computer System Interface (SCSI) adapter, a RAID controller, a SAN adapter, a network adapter, and / or any component providing the processor 206 with access to the database 204.

[0066] In one non-limiting example, the database 204 is configured to store client intent information 218, the first ML model 220, and the one or more second ML models 222and the like. In a non-limiting example, the database 204 is also configured to store the historical conversational interaction dataset that may be used to train one or more ML models, such as the first ML model 220 and the one or more second ML models 222. In one embodiment, the historical conversational interaction dataset may include information related to a plurality of conversational interactions between a plurality of clients and a plurality of agents. Examples of conversational interactions include email conversations, chat conversations, transcribed verbal communication, and the like.

[0067] The processor 206 includes suitable logic, circuitry, and / or interfaces to execute operations for receiving the conversational interaction, generating one or more conversational interaction response templates, and the like. Examples of the processor 206 include, but are not limited to, an Application-Specific Integrated Circuit (ASIC) processor, a Reduced Instruction Set Computing (RISC) processor, a Graphical Processing Unit (GPU), a Complex Instruction Set Computing (CISC) processor, a Field-Programmable Gate Array (FPGA), and the like.

[0068] The memory 208 includes suitable logic, circuitry, and / or interfaces to store a set of computer-readable instructions for performing the various operations described herein. Examples of the memory 208 include a random-access memory (RAM), a read-only memory (ROM), a removable storage drive, a hard disk drive (HDD), and the like. It will be apparent to a person skilled in the art that the scope of the disclosure is not limited to realizing the memory 208 in the server system 200, as described herein. In another embodiment, the memory 208 may be realized in the form of a database server or a cloud storage working in conjunction with the server system 200, without departing from the scope of the present disclosure.

[0069] The processor 206 is operatively coupled to the communication interface 210, such that the processor 206 is capable of communicating with a remote device (i.e., to / from a remote device 224) such as third-party servers or with electronic devices associated with the agents 104 and the clients 106, or communicating with any entity connected to the network 108 (as shown in FIG. 1).

[0070] It is to be noted that the server system 200 as illustrated and hereinafter described is merely illustrative of an apparatus that could benefit from embodiments of the present disclosure and, therefore, should not be taken to limit the scope of the present disclosure. It is noted that the server system 200 may include fewer or more components than those depicted in FIG. 2.

[0071] In one implementation, the processor 206 includes a data receiving module 226, a data pre-processing module 228, an entity extraction module 230, an automatability determination module 232, and a resolution determination module 234. It should be noted that components, described herein, such as the data receiving module 226, the data pre-processing module 228, the entity extraction module 230, the automatability determination module 232, and the resolution determination module 234 can be configured in a variety of ways, including electronic circuitries, digital arithmetic, and logic blocks, and memory systems in combination with software, firmware, and embedded technologies. In various non-limiting examples, the data receiving module 226, the data pre-processing module 228, the entity extraction module 230, the automatability determination module 232, and the resolution determination module 234 may be communicatively coupled with each other and can transfer data between the modules.

[0072] In an embodiment, the data receiving module 226 includes suitable logic and / or interfaces for receiving the conversational interaction initiated by a client (e.g., the client 106A). The received conversational interaction may be provided to the data pre-processing module 228 for further processing.

[0073] In one embodiment, the data pre-processing module 228 includes suitable logic and / or interfaces for extracting a set of relevant sections from the conversational interaction. In a non-limiting implementation, upon receiving the conversational interaction from the client 106A, the conversational interaction is preprocessed for identifying a client intent associated with the conversational interaction. The pre-processing of the conversational interaction includes extracting the set of relevant sections from the conversational interaction. Herein, the set of relevant sections refers to one or more sections of the conversational interaction that may be sufficient or important for identifying the client’s intent. For instance, when the conversational interaction is an email, then the set of relevant sections including a subject line and an email body of the email can be sufficient for identifying the client intent as it is known from predefined email formats that those sections are supposed to include the actual intentions of the email sender.

[0074] In a non-limiting implementation, the extraction of the set of relevant sections from the conversational interaction may be performed using simple rule-based methods or Regular Expressions (Regex) techniques. When such methods are used, simple rules related to a specific pattern of text may be defined by a user such as an administrator (not shown) of the server system 200. Based on the rules defined for a particular pattern, text in the conversationalinteraction may be processed for identifying matching patterns. For example, for identifying the subject line in an email, the subject line pattern such as “Subject:’ followed by “text” may be predefined. Then the data pre-processing module 228 may be configured to process the conversational interaction for identifying the predefined pattern for the subject line. Upon identifying the predefined subject line pattern, the data pre-processing module 228 may extract the subject line from the email. Similarly, the email body may also be extracted from the email. In some embodiments, when the text may be complex, Natural Language Processing (NLP) techniques may be used in place of the rule-based methods and the Regex techniques. In some other embodiments, Regex can be used in combination with NLP techniques. It is to be noted that NLP techniques are powerful for understanding the semantic meaning of text, whereas Regex is often used for pattern-based matching. In the case of using NLP techniques, a pretrained NLP-based model may be used, or predefined language-based libraries may be used for extracting the relevant text from the conversational interaction.

[0075] Upon extracting the set of relevant text, in another embodiment, the data preprocessing module 228 may be configured to generate a first relevant text based, at least in part, on applying a first set of preprocessing operations on the set of relevant sections. In an example scenario, when the query is received via an email, the subject line and the email body are the most relevant sections of the email. Further, the subject line and the email body may be concatenated to identify the client’s intent from the email. Thus, in one embodiment, the first set of preprocessing operations may include a concatenation operation to be performed on the set of relevant sections. As used herein, the term ‘concatenation operation’ refers to an operation of joining multiple strings of characters into a single string. Herein, the characters may include alphanumeric characters, numbers, alphabets, etc. In a non-limiting example, the concatenation operation may be implemented using concatenation operators such as + and & in a preferred pre-processing platform used for the implementation of the same.

[0076] In an embodiment, it may be noted that the methods and systems proposed in the present disclosure can be used in any domain or industry. However, for the sake of explanation, analysis, and performance comparison, the various embodiments of the proposed system are applied in the shipment and logistics industry. However, the same should not be construed as a limitation, and the various embodiments of the present disclosure are applicable in various other industries such as healthcare, retail, media, travel, crime detection, and the like, and the same would be covered within the scope of the present disclosure as well.

[0077] In a specific scenario, in the case of emails, some additional ruleset may bedefined such as emails qualifying the following two criteria may be further passed for identification of the client intent: i. Sub Type = ‘Release Shipping Instructions (SI) / BL / Voyage Charter (VC)’ ii. Reason For Case = ‘B / L Enquiry’ or ‘Release B / L Web Request’

[0078] In some embodiments, the first relevant text may be further pre-processed for the identification of a predefined set of keywords. For example, in a scenario of the client and the agent communicating regarding a shipment, the predefined set of keywords can include ‘swb’, ‘seaway’, ‘bl’, ‘obi’, ‘mbl’, ‘bill’, ‘lading’, ‘waybill’, ‘bls’, ‘swbs’, and the like. One of the rules can also be that the case sensitivity is not applicable.

[0079] In some other embodiments, the data pre-processing module 228 may further be configured to filter out and discard one or more conversational interactions from being automatable at the stage of pre-processing based at least on the above-mentioned rules. For instance, the data pre-processing module 228 may drop the conversational interaction such as the email from the automation consideration if at least the above-mentioned rules don’t qualify. Alternatively, the data pre-processing module 228 may drop the conversational interaction from the automation consideration if the email has mentioned a party name or service charge (SCV) code that is not the same as the account code of the client 106A who raised the request (i.e., sent the email with a query) in the customer service team such as a salesforce team. Upon pre-processing and confirming that the conversational interaction satisfies the above- mentioned rules, the first relevant text may be provided to the entity extraction module 230 and the automatability determination module 232 for further processing by the automatability determination module 232.

[0080] In an embodiment, the automatability determination module 232 includes suitable logic and / or interfaces for generating the client intent information 218 for the conversational interaction based, at least in part, on the first relevant text and the historical conversational interaction dataset. Various AI / ME models may be used by the automatability determination module 232 for generating the client intent information 218. In a specific embodiment, the automatability determination module 232 may be configured to generate the client intent information 218 using the first ME model 220. In an example, the first ML model 220 can be a Large Language Model (LLM), such as a Bidirectional Encoder Representations from Transformers (BERT) model, a Generative Pre-trained Transformer (GPT) model, or the like.

[0081] As may be understood, an LLM is generally trained using a diverse and extensive dataset for language understanding. This dataset covers a wide range of topics, styles, and languages to ensure the model’s generalization ability. However, when the LLM is used for intent identification from a text string, then it may be trained using a labeled dataset having examples of text strings along with their corresponding labels. Herein, the said labeled dataset also should be diverse and cover a wide range of intents that the model is expected to recognize. Thus, the automatability determination module 232 may be configured to train the first ML model 220 to determine the client intent associated with the conversational interaction based on the first relevant text and the historical conversational interaction dataset. In an embodiment, as a labeled dataset is considered for training the first ML model 220, supervised training / learning-based techniques may be preferred for training the first ML model 220. Further, it may be noted that training techniques such as ML models are well-known in the art. Thus, training, validation, and testing stages for the first ML model 220 may be implemented in a way similar to how it may be performed for any Al or ML models and hence these stages are not elaborated herein for the sake of brevity.

[0082] Once the first ML model 220 is trained for intent identification, the first ML model 220 may be used for identifying the client intent associated with the received conversational interaction. Further, the first ML model 220 may generate the corresponding client intent information 218. As described earlier, the client intent information 218 is stored in the database 204 which can be accessed in the future for further processing or any other application.

[0083] In one embodiment, the client intent information 218 may include information related to the client’s intent. In an example scenario, the information may include a description of the client’s intent in text. Herein, it is to be noted that the conversational interaction may be associated with multiple clients. Moreover, in some embodiments, the first ML model 220 can detect multiple client intents in the conversational interaction. Thus, the client intent information can include the information related to the multiple client intents.

[0084] In some additional embodiments, the client intent information 218 can be further refined by training the first ML model 220 to utilize one or more intent-related rulesets such as a first predefined ruleset. Thus, the automatability determination module 232 may further be configured to fine-tune the client intent information 218 based, at least in part, on the first relevant text and the first predefined ruleset. In one embodiment, the automatability determination module 232 may fine-tune the client intent information 218 using the first MLmodel 220. In a non-limiting implementation, along with LLM, an NLP model can also be clubbed with the LLM as the first ML model 220 for fine-tuning the client intent information 218 for better performance.

[0085] For instance, when the query from the client 106 A is received as an email, as a part of the first predefined ruleset, the first ML model 220 may generate responses to the following questions: i. Does the client 106A have problems accessing a requested document in the query on the web? [Yes / No] ii. Is the client 106A asking for any updates to show on the bill? [Yes / No] iii. Is it a request to deliver the bill physically or online? [Physically / Online]

[0086] More specifically, if the response from the first ML model 220 for the delivery of email is ‘Physically’, then since the operation of ‘releasing on destination’ is not automatable, the corresponding email may be dropped from considering it for automation consideration. Further, web errors and amendment emails are also dropped based on the responses of the first ML model 220.

[0087] Further, the conversational interactions that come within the scope of automation can be categorized into one or more automatable process categories. Thus, post intent identification, multiple operations may be performed over the intent i.e., the client intent information 218 to identify the right scope for automation. In one embodiment, the automatability determination module 232 may be configured to perform a set of operations using the one or more second ML models 222. In an example, the one or more second ML models 222 can be a rule-based model, an NLP-based model, a Regex, any other Al or ML model, and / or a combination thereof.

[0088] In a specific non-limiting implementation, the automatability determination module 232 may train the one or more second ML models 222 based at least on the following rules: i. If in the intent it’s determined that the client 106A wants the document either ‘through email’, ‘via email’, or ‘on email’ then drop the email because this cannot be considered under the automation consideration. ii. If the intent states that there is a requirement of a certain bill instead of some other bill, then drop the email on account of the ‘Bill Type Change Request’ . iii. For topic mining, it may be noted that the client intent identified using the first MLmodel 220 is used for process categorization based on the rules. If topic / process category cannot be identified from the first ML model 220, then another ML model such as a KeyBert-based model may be used. In an embodiment, the KeyBert-based model may be used to extract key phrases from the email received from the client 106A. Further, the KeyBert-based model may pass these key phrases as a part of the previously mentioned keywords and perform multiple validations to finally identify the topic (or the process category).

[0089] In one embodiment, the conversational interactions that are associated with multiple client intents may also have to be discarded from the automation consideration. Thus, the automatability determination module 232 may be configured to detect the presence of one or more sub-intents in the client intent information 218.

[0090] In one scenario, detecting the presence of the one or more sub-intents in the client intent information 218 may include detecting two or more sub-intents in the client intent information 218 based, at least in part, on an identification of one or more conjunction texts in the client intent information 218. For example, the query is received via an email, and the client intent information 218 for the email includes the one or more conjunction texts such as ‘AND’ . Then, it may be understood that the client intent information 218 includes multiple client intents. Thus, the automatability determination module 232 may assign a non-automatable process to the conversational interaction based at least on the client intent information 218 being associated with at least two or more sub-intents. Upon classifying the conversational interaction to be a non-automatable process, it may be discarded from further processing.

[0091] In another scenario, detecting the presence of one or more sub-intents in the client intent information 218 may include detecting one sub-intent. Upon detecting only one sub-intent to be available in the client intent information 218, the automatability determination module 232 may be configured to assign a specific automatable process to the conversational interaction based, at least in part, on comparing the client intent information 218 with a set of pre-defined automatable processes. Further, the set of pre-defined automatable processes may be based on the above-listed rules for process categorization. Herein, the term ‘automatable process’ refers to a process that can be automated, i.e., whose steps can be automatically performed by the server system 200 without any intervention from the agent 104A. These processes may refer to processes or tasks that are supposed to be performed to determine a resolution for the query in the conversational interaction received from the client 106A.

[0092] In an embodiment, the entity extraction module 230 includes suitable logic and / or interfaces for extracting a plurality of relevant entities from the conversational interaction based, at least in part, on one or more relevant texts associated with the conversational interaction. In a non-limiting example, the one or more relevant texts may include the first relevant text, the second relevant text, the third relevant text, and the like.

[0093] In a specific embodiment, extracting the plurality of relevant entities may include extracting a first set of relevant entities of the plurality of relevant entities. In a scenario of a shipment being performed between the client 106 A and the employer of the agent 104 A, the first set of relevant entities may include a shipment, number, a customer number, a bill number, and the like. It may be understood that the first set of relevant entities may correspond to entities that are used as identifiers of a product or a service and are essential elements that may be used by the agents 104 or the server system 200 for determining the resolution of the query in the conversational interaction from the client 106A. These entities may need to be mentioned in a resolution email that may be sent to the client 106 A from the agent 104 A upon determination of the resolution for the query received in the email.

[0094] In an embodiment, for extracting the first set of relevant entities, the entity extraction module 230 may be configured to generate the second relevant text based, at least in part, on applying a second set of preprocessing operations on the first relevant text. It is to be noted that, for extracting the first set of relevant entities from the first relevant text, unnecessary noise may have to be eliminated. For example, in the case of the shipment industry, parameters such as the shipment number, customer ID, and other similar parameters may be present in either the subject line, or the email body, and not in any other sections of the email. Also, it should be noted that while processing the conversational interaction, the above- mentioned parameters may be considered as entities and Al-based techniques for extraction of such entities may be considered. Thus, in a non-limiting example, the second set of preprocessing operations may include removing noise, such as salutation, disclaimer, trails of the email, stop words, punctuations, any emails, URLs, numbers, and the like from the first relevant text. Upon removing the unnecessary text from the first relevant text, the second relevant text may be obtained.

[0095] In another embodiment, the entity extraction module 230 may be configured to process the second relevant text to generate the first set of relevant entities based, at least in part, on the second relevant text and a second predefined ruleset. In a non-limiting example, the entity extraction module 230 may process the second relevant text using the one or moresecond ML models 222. As may be understood, the one or more second ML models 222 can be any type of rule -based model, the rules for a pattern associated with the text to be extracted may be defined. Thus, the second predefined ruleset may include rules that define the pattern of the text to be classified as the second relevant text. For example, a shipment number generally includes alphanumeric characters that can include numbers, alphabets, and sometimes special characters as well. Similarly, in another example, the customer code can also include alphanumeric characters. Moreover, such entities have the said characters arranged in a predefined order. Thus, in a non-limiting scenario, the second predefined ruleset can include a rule stating that if a series of alphanumeric characters are detected containing at least two alphabets are followed by a predefined count of random numbers. Then the corresponding entity can be classified in the first set of relevant entities.

[0096] In another specific embodiment, extracting the plurality of relevant entities from the conversational interaction may include extracting a second set of relevant entities of the plurality of relevant entities. In a scenario of a shipment being performed between the client 106 A and the employer of the agent 104 A, the second set of relevant entities may include a sender’s name, a client’s identifier, or the like. It may be understood that the second set of relevant entities may correspond to entities that are required to be mentioned in the resolution email in the salutation part of the resolution email.

[0097] In an embodiment, for extracting the second set of relevant entities, the entity extraction module 230 may be configured to generate the third relevant text based, at least in part, on applying a third set of preprocessing operations on the first relevant text. In a nonlimiting example, the third set of preprocessing operations may include removing unnecessary noise from the first relevant text. For example, in the case of shipment, this step may be performed to obtain the sender’s name and determine that the signature section of the email may be required for further processing. Thus, in this scenario, this relevant text may include a signature section of the email.

[0098] Further, in another embodiment, the entity extraction module 230 may be configured to process the first relevant text to generate an intermediate set of entities based, at least in part, on a third predefined ruleset. In a non-limiting example, the entity extraction module 230 may generate the intermediate set of entities using one or more second ML models 222 such as a Named Entity Recognition (NER) model. Generally, in accordance with the third predefined ruleset, the NER model may be used in scenarios where the entities correspond to names of people, locations, organizations, dates, etc. Thus, it may be noted that the NER modelmay be trained to identify and classify such entities from the text string. In one embodiment, the intermediate set of entities may include a list of names that are proper nouns, i.e., all the names in the email which are names of people, places, organizations, things, or the like.

[0099] In yet another embodiment, the entity extraction module 230 may be configured to compare the third relevant text with the intermediate set of entities to generate the second set of relevant entities. Upon comparison, one or more entities of the intermediate set of entities that match with an entity in the third relevant text may be considered as part of the second set of relevant entities. For example, the third relevant text can be the signature portion of the email, and the intermediate set of entities includes all the proper nouns in the email. The proper nouns can include, the name in the salutation section, email body, and other trail emails. Upon comparing the third relevant text with the intermediate set of entities, a name in the intermediate set of entities matches the name in the signature section is the sender’s name. Hence, the matched name can be used in the resolution email that is supposed to be shared with the client 106A as a reply to the conversational interaction of the client 106A.

[0100] Upon extracting the plurality of relevant entities from the conversational interaction, the data receiving module 226 may be configured to access a list of tasks associated with the specific automatable process from the database 204. As mentioned earlier, each process may be associated with a set of tasks to be performed for determining the resolution for the query of the client 106A received as the conversational interaction. This set of tasks may be predefined, and information related to them may be stored in the database 204. Thus, this information may have to be extracted from the database 204. Further, the set of tasks may have to be performed by the server system 200 to obtain the resolution. The data receiving module 226 may share the information associated with the set of tasks to the resolution determination module 234 for further processing.

[0101] In one embodiment, the resolution determination module 234 includes suitable logic and / or interfaces for generating query resolution information for the conversational interaction based, at least in part, on processing the client intent information for the list of tasks. It should be noted that the list of tasks may include tasks that are specific to a particular process category. For example, in the shipping and logistics industry, the set of tasks may include checking which particular portal has to be visited to check the status of a requested document, checking if the requested document is already generated or not, checking whether it is internally verified, identify an operator of the shipment, and the like. In one embodiment, for processing the client intent information for the list of tasks, the resolution determinationmodule 234 may further be configured to perform each task from the list of tasks to generate the query resolution information for the conversational interaction.

[0102] Once all the tasks are performed and the output of each task performed is positive, a relevant resolution including a positive response may be generated. Alternatively, a resolution including a negative response may be generated if the outcome of performing any of the set of tasks is negative. For example, if the request was for the generation of an invoice. In this scenario, if the invoice was already generated, the resolution for the query could be ‘Yes, the invoice is generated and shared across’. On the other hand, if an invoice was not generated, then the resolution determination module 234 may check for the possibility of generating the said invoice. If it is not possible, a message may be passed to the client 106A as a reply to the query stating that: ‘No, the invoice is not generated’.

[0103] In another embodiment, the resolution determination module 234 may further be configured to generate one or more conversational interaction response templates for the conversational interaction based, at least in part, on the plurality of relevant entities, and the query resolution information. Herein, the one or more conversational interaction response templates may be generated for the agent 104 A as recommendations. The agent 104 A may choose the most appropriate template from the one or more conversational interaction response templates and share the same with the client 106A as a response conversational interaction to the received conversational interaction from the client 106A.

[0104] FIG. 3 is a flowchart 300 depicting a process of generating Al resolution templates for an agent (e.g., the agent 104A) in response to a client’s query email from a client (e.g., the client 106A), in accordance with an embodiment of the present disclosure. In an example scenario, the client 106A may visit a shipment assistance platform on a mobile device of the client 106A. Upon visiting the platform, the client 106A places an order for transporting a shipping container full of various goods from a first location to a second location. In this scenario, it is assumed that the shipping container has to cross international borders before reaching the desired location. Therefore, it may be transported via a cargo ship by a shipping company.

[0105] While placing the order for shipment through the online platform associated with the shipping company i.e., the shipment assistance platform, the client 106A might be given multiple options. These options may be shown to client on the shipment assistance platform installed on their electronic device through a GUI. The said options may be related tooptions for setting payment modes, setting payment timings, filling in shipment details, providing shipment instructions, editing personal details, and the like. Suppose the client 106A chooses to make the payment once the product reaches its destination location. Once that happens, the client 106A may send an email either using login credentials on the shipment assistance platform or a personal emailing account, such as that of Google®, Yahoo®, Hotmail®, or the like.

[0106] On the other hand, the shipping company might have hired a team of agents such as the agents 104 to resolve the queries of the clients 106. Thus, the agents 104 may also be registered on the shipment assistance platform with administrative privileges. That may give the agents 104 access to the queries received from the clients 106 through admin login credentials on electronic devices that are associated with the agents 104 individually. The agent 104 A may receive the email from the client 106 A and hence is responsible for resolving the issues that the client 106A is facing and closing the case.

[0107] As may be understood, the container shipping process is associated with the generation and management of one or more documents that can be recorded as proof and can be used in the future. In one embodiment, the one or more documents may include a commercial invoice, a B / L, an import license, an arrival notice, a letter of credit, and the like for importing products. In another embodiment, the one or more documents may further include the B / L, a packing list, a commercial invoice, a letter of credit, a certificate of origin, declarations, certificates, or the like for exporting products.

[0108] It should be noted that at the end of the agents 104, the shipment assistance platform may be communicably coupled with the server system 200. Thus, features facilitated by the server system 200 may be available for the benefit of the agents 104. So, when the client 106A sends the query email to the agent 104A, the server system 200 may receive the query email as well. The steps implemented by the server system 200 are depicted in FIG. 3.

[0109] Upon receiving the query email, the server system 200 extracts the set of relevant sections such as ‘subject’ and ‘email body’ from the query email and concatenates them. Upon concatenation, the server system 200 may generate the first relevant text which is a concatenated result. The concatenated result may include subject + email body (see, 302) as input to the first ML model 220 such as the LLM 304. For example, the query email can be “SUBJECT OF EMAIL: REQ FOR SEAWAY BL / BL NO: 12*****89 / 22*****23 / !3*****23 / 4i*****34. EMAIL FROM CUSTOMER: DEAR ABCDEF TEAM,ATTACHED PAYMENT RECEIPT FOR THE SUBJECT SHIPMENT. KINDLY POST THE “SEAWAY BL” AT EARLIEST. THANKS & REGARDS, AAAAA.C, SR. EXECUTIVE - DOCUMENTATION, ABCABC LOGISTICS PVT LTD, NO.123, T2 & T3, THIRD FLOOR, XYZ MALL, DEFDEF ROAD, XXXXX, YYYYYY- 123456, MOBILE:+9I 98 ** **** ](j TEL NO: o**-*****67”.

[0110] When this email is received by the agent 104A, the server system 200 access the same as an input. Further, the first relevant text 302 is provided to the LLM 304 for performing the operations of intent identification (see, 306), and process categorization (see, 308). The operation of the intent identification 306 results in providing a client intent associated with the query email in the form of the client intent information. In addition, the client intent information is fine-tuned based on the first predefined ruleset in which the LLM 304 is trained to answer a set of questions as mentioned earlier. Upon performing the intent identification 306, the intent identified for the received query email can be “REQUEST TO POST SEAWAY BL FOR THE MENTIONED SHIPMENT”. Further, in a non-limiting implementation, an output of the LLM 304 may be similar to a table as shown in FIG. 4A (see, 402). It is to be noted that the table 402 shown in FIG. 4A may be the format in which the output of the LLM 304 may be displayed on a UI of any electronic device of at least one of the client 106A and the agent 104A.

[0111] Referring to FIG. 4A (see, 400), it may be observed that a first column 404 in the table 402 shows the query email under the title of ‘combined text’, and a second column 406 shows the client intent under the title of ‘Intent’. Further, the third, fourth, and fifth columns (see, 408, 410, and 412, respectively) show answers generated by the LLM 304 for questions ‘web access issue?’, ‘amendment query?’, and ‘require physical query?’. It may be observed that the answer for each of these questions is ‘NO’. Thus, upon identification of the client intent and upon obtaining the answers to questions to be a ‘NO’, then the process moves forward to the step of process categorization 308.

[0112] Referring back to FIG. 3, in the step of process categorization 308, the server system 200 may further process the first relevant text 302 to determine whether the client intent associated with the first relevant text 302 can be categorized as an automatable process (see, 310). The server system 200 may determine the automatability associated with the first relevant text 302, based on a comparison with the set of pre-defined automatable processes and a predefined set of rules explained earlier in the present disclosure. Moreover, in another embodiment, the server system 200 may use the one or more second ML models 222 forperforming the process categorization 308.

[0113] In some embodiments, the server system 200 may train the one or more second ML models 222 based on the set of rules that may also include a set of keywords. The set of rules may be as follows: i. If the keywords from the first two lists below appear in the same sentence in the client intent, then it may be assigned with a process category such as ‘Release SWB’, provided none of the words in excluded list swb are present:1. identification_keys_swb_l = ['swb', 'seaway', 'seawaybill', 'seawaybl', 'waybill', 'swbbl', 'express_bf, 'swbs', 'get_swb', 'original_swb'] .2. identification_keys_swb_2 = ['release', 'provide', 'post', 'upload', 'arrange', 'publish', 'issue', 'request', 'get_swb'].3. excluded_list_swb = ['send', 'edo', 'do', 'freight', 'rated', 'unrated', 'amend', 'revise', 'cdo', 'rates', 'obi', 'mbl', 'vc', 'verify_copy', 'draff, 'charge', 'not_issue', 'update', 'change', 'non-rated', 'not_issue' 'withdraw', 'print', 'destination', 'collection', 'of_issue', 'status', 'confirm', 'copy', 'counter', 'collect', 'unfreighted', 'non_rated', 'rectify', 'consignee', 'reupload', 're-posf, 'original']. ii. If keywords from the first two lists below appear in the same sentence in the client intent, then it may be assigned with a process category such as ‘Release OBL\MBL’, provided none of the words in excluded list obl are present:1. identification_keys_obl_l = ['bl', 'obi', 'mbl', 'bof, 'bill', 'bills', 'bls', 'get_ob'] .2. identification_keys_obl_2 = ['release', 'provide', 'post', 'upload', 'arrange', 'publish', 'issue', 'request', 'final', 'get_obl'].3. excluded_list_obl = ['send', 'destination', 'rated', 'unrated', 'edo', 'do', 'amend', 'revise', 'print', 'cdo', 'charge', 'not_issue', 'freight', 'unfreighted', 'change', 'amendment', 'print', 'update', 'collect', 'split', 'pod', 'surrender', 'freight', 'address', 'instead_of , 'approve', 'switch_bf, 'courier', 'counter', 'withdraw', 'pick_up', 'collection', 'of_issue', 'status', 'confirm', 'copy', 'collect', 'slave', 'confirmation', 'overdue', 'transport_bf, 'unfreighted', 'non_rated', 'rectify', 'consignee', 'reupload', 're-post'].

[0114] In one embodiment, upon process categorization 308, if the query email is classified as non-automatable, then the case may be dropped (see, 312). Alternatively, if the query email is classified to be automatable, then the process moves to the step of extractingthe plurality of entities (see, 314). The output of performing the step of process categorization 308 can be similar to a table (see, 414) shown in FIG. 4A.

[0115] Again, referring to FIG. 4A, in a non-limiting example, the table 414 shows two different intents that are assigned with certain automatable process categories. One of the client intents includes ‘THE CUSTOMER IS REQUESTING THE UPLOAD OF FINAL SWB FOR BOOKING #11***** 12’ as shown in column 416 under a title of combined text is assigned with a process category of ‘RELEASE SEAWAY BILL’ shown in column 418. This classification may be performed based on the above-mentioned rules. Herein, the term ‘combined text’ is used to refer to the client’s intent. Similarly, another example of the client intent is ‘THE CUSTOMER INTENT IS TO RELEASE THE OBL ON AN URGENT BASIS’ which is assigned with a process category of ‘RELEASE OBL\MBL’ .

[0116] Further, referring back to FIG. 3, as may be understood, the plurality of entities includes the first relevant text 302, the second relevant text, and the third relevant text. Herein, the second relevant text may be obtained upon performing the step of preprocessing (see, 316) on the first relevant text 302. Similarly, the third relevant text may be obtained upon performing a preprocessing operation such as using an NER system (see, 318). In an embodiment, the NER system 318 may use the NER model for preprocessing the first relevant text 302. The output of the NER system 318 is the sender’s name (see, 320).

[0117] Further, upon identification of the process category, extraction of the entities, and extraction of the sender’s name, the server system 200 may perform a respective set of tasks. The tasks may be performed to get the resolution for the query email. In a non-limiting example of a GCSS portal (see, 322), for the process category of ‘RELEASE SEAWAY BILL’, the set of tasks may include ‘TASK 1’, ‘TASK 2’, ‘TASK 3’, ‘TASK 4’, and ‘TASK 5’ as shown in FIG. 3. Each task may have to be performed and checked whether the for the client intent, the task passes to the client’s requirements.

[0118] More specifically, the set of tasks may be as follows: i. For GCSS automation, as mentioned above, there are in total of 5 steps involved. Before implementing the tasks described earlier, an operator of the shipment may also have to be identified, as it is being used in the GCSS API call URLs as a carrier. In a non-limiting example, the mapping is as follows:1. if operator in ['MSL', 'MAEU']: carrier = 'MAEU'2. elif operator in ['SGL', 'SEJJ']: carrier = 'SEJT3. elif operator in ['SEA', 'SEAU']: carrier = 'SEAU'4. elif operator in ['MCC, 'MCPU']: carrier = 'MCPU'5. elif operator in ['HSU', 'SUDU']: carrier = 'SUDU'6. elif operator in ['MLL', 'MAEI']: carrier = ‘MAEI’ ii. TASK 1: Check if a vessel has departed or not and that can be recognized basis of a particular OPS task where it is required to identify whether the pickup city is ‘OUT’ or not. iii. Identify whether the Bill Type and identified topic are the same. For example, for SWB bill type should be ‘TPDoc Sea Waybill Shipped’ and for OBL\MBL it should be ‘TPDoc Negotiable Shipped'. iv. TASK 3: Check if the Approve Freight Release (AFR) Documentation task is closed. v. TASK 4: Check if Issue Original Transport Document (IOTD) is closed. vi. TASK 5: Match whether the account sending the email to a shipping service provider, its SCV code is tagged as Transport Document Party (TDR Party) in the GCSS party task. If any other party is tagged as TDR other than the account sending the mail, then consider it as a check failed. vii. When all the tasks are completed and all checks passed, the server system 200 may then formulate a response email informing the client 106 A that the respective document being asked, has been uploaded on the website. viii. In case of multiple shipments, consolidated status of all, in one single email may be provided.

[0119] In one embodiment, the server system 200 may further be configured to generate the one or more conversational interaction response templates which is indicated by step 324 in FIG. 3. The one or more conversational interaction response templates can be referred to as ‘Al templates’, ‘Al resolution templates’, or ‘templates’ (see, 326). For generating the templates 326, the server system 200 may also consider the third relevant text such as the sender’s name which may be used in the salutation section of the query email. Upon performing this step, the server system 200 provides the corresponding one or more Al templates 326 to the agent 104A as a recommendation for the agent 104A. Eater, the client 106A may choose one of the recommendations and draft a response email to be sent to the client 106A.

[0120] In one embodiment, referring to FIG. 4B (see, 460), it may be understood thatthree possible recommended Al templates 462, 464, and 466 may be generated by the server system 200 in an example scenario. Herein, the recommended Al templates 462-466 are examples of the Al templates 324 described in FIG. 3. The first recommended template 462 is generated in a way shown in FIG. 4B, based on the sender’s name extracted from an email signature using the NER model, the shipment number extracted from the client query email, and a TDR party name and SCV code extracted from the GCSS portal.

[0121] It may be understood that these templates have been generated based on historical replies. Additionally, the templates are masked at positions where information about new clients will be inserted for future cases. Further, in an embodiment, two more types of templates may be recommended such as a second recommended template 464, and a third recommended template 466 based on the response of one or more tasks performed from the set of tasks that are performed for determining the resolution for the query. For instance, the second recommended template 464 may be generated, if the TDR party is the same as the account sending mail. The third recommended template 466 may be generated, if the TDR party is anyone else other than the account sending the email.

[0122] FIG. 5 illustrates a flow diagram of a method of generating automated replies for conversational interactions, in accordance with an embodiment of the present disclosure. The method 500 depicted in the flow diagram may be executed by, for example, the server system 200. The sequence of operations of the method 500 may not be necessarily executed in the same order as they are presented. Further, one or more operations may be grouped and performed in the form of a single step, or one operation may have several sub-steps that may be performed in parallel or in a sequential manner. Operations of the method 500, and combinations of operations in the method 500 may be implemented by, for example, hardware, firmware, a processor, circuitry, and / or a different device associated with the execution of software that includes one or more computer program instructions. The plurality of operations is depicted in the process flow of the method 500. The process flow starts at operation 502.

[0123] At 502, the method 500 includes receiving, from a first Machine Learning (ML) model (e.g., the first ML model 220), client intent information (e.g., the client intent information 218) for a conversational interaction initiated by a client (e.g., the client 106A).

[0124] At 504, the method 500 includes performing, by one or more second ML models (e.g., the one or more second ML models 222), a set of operations. In one embodiment, the set of operations may include operations 504A-504F.

[0125] At 504A, the method 500 includes detecting the presence of one or more subintents in the client intent information 218.

[0126] At 504B, the method 500 includes upon detecting one sub-intent, assigning a specific automatable process to the conversational interaction based, at least in part, on comparing the client intent information 218 with a set of pre-defined automatable processes.

[0127] At 504C, the method 500 includes extracting a plurality of relevant entities from the conversational interaction based, at least in part, on one or more relevant texts associated with the conversational interaction.

[0128] At 504D, the method 500 includes accessing a list of tasks associated with the specific automatable process from a database (e.g., the database 204).

[0129] At 504E, the method 500 includes generating query resolution information for the conversational interaction based, at least in part, on processing the client intent information 218 for the list of tasks.

[0130] At 504F, the method 500 includes generating one or more conversational interaction response templates (e.g., 462, 464, and 466) for the conversational interaction based, at least in part, on the plurality of relevant entities, and the query resolution information.

[0131] The disclosed method with reference to FIG. 5, or one or more operations of the server system 200 may be implemented using software including computer-executable instructions stored on one or more computer-readable media (e.g., non-transitory computer- readable media, such as one or more optical media discs, volatile memory components (e.g., DRAM or SRAM), or nonvolatile memory or storage components (e.g., hard drives or solid- state nonvolatile memory components, such as Flash memory components) and executed on a computer (e.g., any suitable computer, such as a laptop computer, netbook, Web book, tablet computing device, smartphone, or other mobile computing devices). Such software may be executed, for example, on a single local computer or in a network environment (e.g., via the Internet, a wide-area network, a local-area network, a remote web-based server, a client-server network (such as a cloud computing network), or other such networks) using one or more network computers.

[0132] Additionally, any of the intermediate or final data created and used during the implementation of the disclosed methods or systems may also be stored on one or more computer-readable media (e.g., non-transitory computer-readable media) and are considered to be within the scope of the disclosed technology. Furthermore, any of the software-basedembodiments may be uploaded, downloaded, or remotely accessed through a suitable communication means. Such suitable communication means include, for example, the Internet, the World Wide Web (WWW), an intranet, software applications, cable (including fiber optic cable), magnetic communications, electromagnetic communications (including RF, microwave, and infrared communications), electronic communications, or other such communication means.

[0133] Although the invention has been described with reference to specific exemplary embodiments, it is noted that various modifications and changes may be made to these embodiments without departing from the broad scope of the invention. For example, the various operations, blocks, etc., described herein may be enabled and operated using hardware circuitry (for example, Complementary Metal Oxide Semiconductor (CMOS) based logic circuitry), firmware, software, and / or any combination of hardware, firmware, and / or software (for example, embodied in a machine-readable medium). For example, the apparatuses and methods may be embodied using transistors, logic gates, and electrical circuits (for example, Application Specific Integrated Circuit (ASIC) circuitry and / or Digital Signal Processor (DSP) circuitry).

[0134] Particularly, the server system 200 and its various components may be enabled using software and / or using transistors, logic gates, and electrical circuits (for example, integrated circuit circuitry such as ASIC circuitry). Various embodiments of the invention may include one or more computer programs stored or otherwise embodied on a computer-readable medium, wherein the computer programs are configured to cause the processor or the computer to perform one or more operations. A computer-readable medium storing, embodying, or encoded with a computer program, or similar language, may be embodied as a tangible data storage device storing one or more software programs that are configured to cause the processor or computer to perform one or more operations. Such operations may be, for example, any of the steps or operations described herein. In some embodiments, the computer programs may be stored and provided to a computer using any type of non-transitory computer- readable media. Non-transitory computer-readable media includes any type of tangible storage media.

[0135] Examples of non-transitory computer-readable media include magnetic storage media (such as floppy disks, magnetic tapes, hard disk drives, etc.), optical magnetic storage media (e.g. magneto-optical disks), Compact Disc Read-Only Memory (CD-ROM ), Compact Disc Recordable (CD-R), compact disc rewritable (CD-R / W), Digital Versatile Disc (DVD),BLU-RAY® Disc (BD), and semiconductor memories (such as mask ROM, programmable ROM (PROM), (erasable PROM), flash memory, Random Access Memory (RAM), etc.). Additionally, a tangible data storage device may be embodied as one or more volatile memory devices, one or more non-volatile memory devices, and / or a combination of one or more volatile memory devices and non-volatile memory devices. In some embodiments, the computer programs may be provided to a computer using any type of transitory computer- readable media. Examples of transitory computer-readable media include electric signals, optical signals, and electromagnetic waves. Transitory computer-readable media can provide the program to a computer via a wired communication line (e.g., electric wires, and optical fibers) or a wireless communication line.

[0136] Various embodiments of the invention, as discussed above, may be practiced with steps and / or operations in a different order, and / or with hardware elements in configurations, which are different than those which are disclosed. Therefore, although the invention has been described based on these exemplary embodiments, it is noted that certain modifications, variations, and alternative constructions may be apparent and well within the scope of the invention.

[0137] Although various exemplary embodiments of the invention are described herein in a language specific to structural features and / or methodological acts, the subject matter defined in the appended claims is not necessarily limited to the specific features or acts described above. Rather, the specific features and acts described above are disclosed as exemplary forms of implementing the claims.

Claims

CLAIMS1. A computer- implemented method, comprising: receiving, from a first Machine Learning (ML) model, client intent information for a conversational interaction initiated by a client; and performing, by one or more second ML models, a set of operations comprising: detecting a presence of one or more sub-intents in the client intent information; upon detecting one sub-intent, assigning a specific automatable process to the conversational interaction based, at least in part, on comparing the client intent information with a set of pre-defined automatable processes; extracting a plurality of relevant entities from the conversational interaction based, at least in part, on one or more relevant texts associated with the conversational interaction; accessing a list of tasks associated with the specific automatable process from a database; generating query resolution information for the conversational interaction based, at least in part, on processing the client intent information for the list of tasks; and generating one or more conversational interaction response templates for the conversational interaction based, at least in part, on the plurality of relevant entities, and the query resolution information.

2. The computer- implemented method as claimed in claim 1, wherein receiving the client intent information comprises: receiving the conversational interaction initiated by the client; extracting a set of relevant sections from the conversational interaction; generating a first relevant text of the one or more relevant texts based, at least in part, on applying a first set of preprocessing operations on the set of relevant sections; and generating, by the first ML model, the client intent information for the conversational interaction based, at least in part, on the first relevant text and a historical conversational interaction dataset.

3. The computer- implemented method as claimed in any of claim 1 and claim 2, furthercomprising: fine-tuning, by the first ML model, the client intent information based, at least in part, on the first relevant text and a first predefined ruleset.

4. The computer- implemented method as claimed in claim 2, wherein extracting the plurality of relevant entities comprises: extracting a first set of relevant entities of the plurality of relevant entities based, at least in part, on performing: generating a second relevant text of the one or more relevant texts based, at least in part, on applying a second set of preprocessing operations on the first relevant text; and processing, by the one or more second ML models, the second relevant text to generate the first set of relevant entities based, at least in part, on the second relevant text and a second predefined ruleset.

5. The computer- implemented method as claimed in claim 2, wherein extracting the plurality of relevant entities from the conversational interaction comprises: extracting a second set of relevant entities of the plurality of relevant entities based, at least in part, on performing: generating a third relevant text of the one or more relevant texts based, at least in part, on applying a third set of preprocessing operations on the first relevant text; processing, via the one or more second ML models, the first relevant text to generate an intermediate set of entities based, at least in part, on a third predefined ruleset; and comparing the third relevant text with the intermediate set of entities to generate the second set of relevant entities.

6. The computer-implemented method as claimed in claim 1, wherein processing the client intent information for the list of tasks comprises: performing each task from the list of tasks to generate the query resolution information for the conversational interaction.

7. The computer-implemented method as claimed in claim 1, wherein detecting the presence of the one or more sub-intents in the client intent information comprises:detecting two or more sub-intents in the client intent information based, at least in part, on an identification of one or more conjunction texts in the client intent information.

8. The computer-implemented method as claimed in claim 1, further comprising: assigning a non- automatable process to the conversational interaction based at least on the client intent information being associated with at least two or more sub-intents.

9. A server system, comprising: a communication interface; a memory configured to store instructions; and a processor in communication with the communication interface and the memory, the processor configured to execute the instructions stored in the memory and thereby cause the server system to perform at least in part to: receive, from a first Machine Learning (ML) model, client intent information for a conversational interaction initiated by a client; and perform, by one or more second ML models, a set of operations comprising: detect a presence of one or more sub-intents in the client intent information; upon detecting one sub-intent, assign a specific automatable process to the conversational interaction based, at least in part, on comparing the client intent information with a set of pre-defined automatable processes; extract a plurality of relevant entities from the conversational interaction based, at least in part, on one or more relevant texts associated with the conversational interaction; access a list of tasks associated with the specific automatable process from a database; generate query resolution information for the conversational interaction based, at least in part, on processing the client intent information for the list of tasks; and generate one or more conversational interaction response templates for the conversational interaction based, at least in part, on the plurality of relevant entities, and the query resolution information.

10. The server system as claimed in claim 9, wherein for receiving the client intent information, the server system is caused, at least in part, to:receive the conversational interaction initiated by the client; extract a set of relevant sections from the conversational interaction; generate a first relevant text of the one or more relevant texts based, at least in part, on applying a first set of preprocessing operations on the set of relevant sections; generate, by the first ML model, the client intent information for the conversational interaction based, at least in part, on the first relevant text and a historical conversational interaction dataset; and fine-tune, by the first ML model, the client intent information based, at least in part, on the first relevant text and a first predefined ruleset.

11. The server system as claimed in claim 10, wherein for extracting the plurality of relevant entities, the server system is caused, at least in part, to: extract a first set of relevant entities of the plurality of relevant entities based, at least in part, on performing: generating a second relevant text of the one or more relevant texts based, at least in part, on applying a second set of preprocessing operations on the first relevant text; and processing, by the one or more second ML models, the second relevant text to generate the first set of relevant entities based, at least in part, on the second relevant text and a second predefined ruleset.

12. The server system as claimed in claim 10, wherein for extracting the plurality of relevant entities, the server system is caused, at least in part, to: extract a second set of relevant entities of the plurality of relevant entities based, at least in part, on performing: generating a third relevant text of the one or more relevant texts based, at least in part, on applying a third set of preprocessing operations on the first relevant text; processing, via the one or more second ML models, the first relevant text to generate an intermediate set of entities based, at least in part, on a third predefined ruleset; and comparing the third relevant text with the intermediate set of entities to generate the second set of relevant entities.

13. The server system as claimed in claim 9, wherein for processing the client intentinformation for the list of tasks, the server system is caused, at least in part, to: perform each task from the list of tasks to generate the query resolution information for the conversational interaction.

14. The server system as claimed in claim 9, wherein for detecting the presence of the one or more sub-intents in the client intent information, the server system is caused, at least in part, to: detect two or more sub-intents in the client intent information based, at least in part, on an identification of one or more conjunction texts in the client intent information; and upon detecting at least two or more sub-intents, assign a non-automatable process to the conversational interaction based at least on the client intent information associated with the corresponding two or more sub-intents.

15. A non-transitory computer-readable storage medium comprising computer-executable instructions that, when executed by at least a processor of a server system, cause the server system to perform a method comprising: receiving, from a first Machine Learning (ML) model, client intent information for a conversational interaction initiated by a client; and performing, by one or more second ML models, a set of operations comprising: detecting a presence of one or more sub-intents in the client intent information; upon detecting one sub-intent, assigning a specific automatable process to the conversational interaction based, at least in part, on comparing the client intent information with a set of pre-defined automatable processes; extracting a plurality of relevant entities from the conversational interaction based, at least in part, on one or more relevant texts associated with the conversational interaction; accessing a list of tasks associated with the specific automatable process from a database; generating query resolution information for the conversational interaction based, at least in part, on processing the client intent information for the list of tasks; and generating one or more conversational interaction response templates for the conversational interaction based, at least in part, on the plurality of relevant entities, and the query resolution information.

Images