Deliver Higher Engagement
with human-like conversations
with human-like conversations
Dialog-turn Management
Users often fail to follow a task to its logical conclusion before initiating another, which means that the assistant must guide the conversation in the new direction and provide results – without losing the original context of the conversation.
The intelligent virtual assistants (IVAs) developed on the Kore.ai XO Platform are capable of handling almost all aspects of human conversations, such as interruptions, digressions and clarifications, among others. Furthermore, you have complete control over defining the dialog turn and context-switching experience for users.
Human conversations are twisted and communicate multiple tasks before the logical end of the previous task. The platform provides you granular control with “hold and resume” functionality to handle such scenarios. It allows you to control context switching rules, behaviors and configure for how the assistants behave when intent resumes –
Handling dynamic conversations and switching between intents and entities or combining multiple intents requires robust intelligence.
The platform’s intelligence engine captures and stores all intents and unattended interruptions to maintain the context, thus, offering free-flow conversations to users. It also performs tasks from a list of identified follow-up intents or gives you the ability to trains and refine these flows.
Deliver human-like conversations to your users. The platform’s intelligence engine breaks conversations down to their essence and identifies and follows up on multiple action items or intents in a single message. The assistant can then execute tasks sequentially and logically.
The Kore.ai Experience Optimization Platform is the only one today that can handle multiple intents at the dialog management layer.
Virtual assistants built on the platform allow users to amend entity values at any point in a conversation. The assistant understands the context of the request, identifies the entity to be amended, and then either clarifies the action to be taken or takes action directly – all without writing a single line of code.
The custom small talk feature extends the virtual assistant’s built-in small talk abilities, build a personality and train the assistant to conduct casual conversations. It provides an easy-to-use interface for quickly generating multiple scenarios by providing the user and assistant utterances. The Custom Small Talk UI Editor can be used to build nested conversations to answer follow-up questions and context-specific responses.
Context Management
Effective context management enables virtual assistants to interact with users more quickly and intuitively; it means less robotic and scripted. Contextual data helps assistants complete tasks faster and creates a more natural, human-like back and forth conversations. Personalize experiences to gain higher user satisfaction.
The platform allows you to capture and reuse contextual data for a large variety of scenarios to create more complex use cases and redefine the user experience.
Manage data and contextual details at the framework level – data that can come from user inputs, API responses, sessions, and more – with little coding required. It enables you to create context-aware assistants to follow conversation history, harvest and harness information from third-party systems, and accurately predict and populate appropriate tasks or responses.
When you create tasks, you can access session variables provided by the XO Platform or custom variables you define and the context that defines the variable’s scope.
Store and retrieve data for various use cases, such as personalized greetings, determining where prior conversations left off, and pre-filling order details or recommendations.
You determine how long these variables are stored and accessible by your assistants – so you can create more intuitive conversations with users on your terms.
Humans tend to switch back and forth between intents. IVAs developed on the Kore.ai XO Platform handle these situations (digressions) efficiently.
The platform also offers provisions to configure the conditions to enable or disable context switches; also add conditional exceptions between tasks with the ability to pass contextual data between them. The platform handles virtually all complex and diverse content-switching scenarios effectively.
Create virtual assistants that remember user inputs and answers – and automatically modify their behavior based on sentiment and context.
Connect to your backend business systems to pull and store the necessary data– otherwise often static, contained within a single location, and rarely remembered in context. The platform allows you to create dynamic conversational experiences that can customize, categorize, and apply contextual information as you see fit.
What: Individual user information and preferences can be stored and used in varying tasks. This information can be shared with all assistants during user interaction.
Why: Transactions and interactions are made shorter and easier, as users no longer have to provide the same basic information repeatedly.
Example: Information like a user’s home address, payment information, airline seat, or baggage preference can be stored and remembered for future interactions.
What: Information representing company-wide rules and standards for all users and assistants.
Why: It allows developers to ensure that assistants keep company requirements in context and enforce them as needed.
Example: A company travel or expense policy could be applied that overrides an employee’s preferences when using a travel assistant to book a flight or hotel.
What: Information shared by the user that the assistant must remember during a session.
Why: Avoids users from repeating relevant information to complete multiple tasks in a single session.
Example: A bank customer says, “how much do I have in checking?” and then asks to “transfer $500 to savings.” The assistant could recognize “checking” as session context and thus make the transfer without asking the user, “from which account?”.
What: Information shared by the user that the assistant must remember during a session.
Why: Keeps the user from having to repeat information that’s relevant for completing multiple tasks in a single session.
Example: A bank customer says, “how much do I have in checking?” and then asks to “transfer $500 to savings.” The assistant could recognize “checking” as session context and thus make the transfer without asking the user, “from which account?”.
Sentiment Management
The platform offers sentiment management capabilities. It enables virtual assistants to ‘understand’ the user’s emotions by deciphering verbal and sentence structure clues and parses user utterances for specific words, phrases, and modifiers.
Assistants can handle the happier customers themselves while transferring the unhappier customers to human agents.
Emotions influence people’s choices, how they interact with others and their perception of the world around them – including your brand. The platform analyzes user utterances and classifies them based on detected levels of anger, disgust, fear, sadness, joy, and positivity.
However, emotions are not mutually exclusive and usually experience along a gradient. Our tone algorithm and platform intelligence accounts for this reality and uncovers by scoring multiple emotions in a single utterance. For example, an input could yield a high score for joy but a mild score for sadness at the same time.
The platform’s NL engine scores and ranks sentiment on a scale of -3 to 3, based on the intensity of the emotions detected. This number represents how high or how little the user input corresponds to the six core emotions.
It provides sentiment analysis results as context object variables at two different levels; tone emotions and scores for the current conversation and average tone emotions and scores for the entire session.
Learn how Kore.ai products and solutions can drive ROI, reduce expenses and increase revenue.