Discover the Power of Kore.ai Platform in 20 Minutes

Discover the Power of Kore.ai Platform in 20 Minutes

Table of Contents:

1. Introduction
2. Designing Conversations using Conversations
3. Building the Virtual Assistant
4. Using the Task Execution Framework
5. Digital Views and UI Forms
6. Training the Virtual Assistant
7. Testing the Virtual Assistant
8. Publishing the Virtual Assistant
9. Configuring the Virtual Assistant
10. Analyzing Performance and User Interactions
11. Conclusion

Introduction The Core AI Platform Overview

• Walkthrough of the Virtual Assistant Life Cycle
• Core AI Platform Components and Capabilities
• Benefits and Use Cases of Core Virtual Assistant Platform

Designing Conversations using Conversations

• Overview of the Storyboard Feature
• Creating Scenes for Different Use Cases
• Designing Conversations within Scenes
• Sharing and Reviewing Conversation Designs

Building the Virtual Assistant

• Overview of the Task Execution Framework
• Creating Dialogue Tasks for Complex Conversational Experiences
• Configuring Tasks using Nodes and Interconnections
• Creating Multiple Dialogue Tasks for Different Use Cases

Using the Task Execution Framework

• Enabling Different Types of Tasks
• Configuring Entity Nodes and Message Nodes
• Adding JavaScript with Script Nodes
• Making API Calls with Service Nodes

Digital Views and UI Forms

• Building Interactive Components with Digital Views
• Using Panels, Widgets, and Charts for User Interaction
• Creating UI Forms for Capturing User Details
• Customizing and Configuring UI Forms

Training the Virtual Assistant

• Introduction to NLP Engines
• Using Machine Learning Engine for Training
• Training the Fundamental Meaning Engine
• Utilizing Ontology-Based Knowledge Graph Model
• Configuring Interruption Handling and Sentiment Analysis

Testing the Virtual Assistant

• Introduction to Utterance Testing and Batch Testing
• Testing Virtual Assistants with Various User Inputs
• Analyzing Matched Intents and Failed Tasks
• Uploading Test Suites for Batch Testing

Publishing the Virtual Assistant

• Enabling Channels for User Connections
• Integrating with Popular Messaging and Voice Channels
• Verifying Bot Settings and Languages
• Publishing Bot for End Users

Configuring the Virtual Assistant

• Overview of Config Settings and Bot Management
• Configuring General Settings and PII Redaction
• Defining Authorization Profiles and Variables
• Sharing, Exporting, Importing Bots, and Change Logs

Analyzing Performance and User Interactions

• Understanding Conversation Flows and User Journeys
• Analyzing Intents Found and Not Found
• Monitoring Failed Tasks and API Performance
• Utilizing Usage Metrics and Custom Dashboards

Conclusion

• Recap of Core AI Platform Features
• Significance of Low Code No Code Platform
• Benefits of Core AI for Building Virtual Assistants

Introduction

The core AI platform provides enterprises with the necessary components and capabilities to build sophisticated virtual assistants. Offering a unique Blend of conversational chat interface and digital user experience, the platform empowers organizations to design, build, test, and deploy AI-powered virtual assistants. Developed on the core AI virtual system platform, these intelligent virtual assistants can satisfy complex enterprise use cases, accelerate time to market, and exceed customer expectations. In this article, we will Delve into the different stages of a virtual assistant's life cycle using the core platform.

Designing Conversations using Conversations

The conversation designer in the core AI platform offers a simple and natural way to design virtual assistants. With the storyboard feature, conversations can be captured to represent each use case identified for the bot. Within each scene, an entire conversation can be designed, laying out a series of bot and user conversations. The conversation design can be previewed to see how the bot will look after it gets published. Additionally, the design can be shared with stakeholders for review and approval before building the bot.

Building the Virtual Assistant

Once the conversation is designed, the next step is to build the virtual assistant. The task execution framework provides a powerful tool to build different tasks for the virtual assistant. Dialogue tasks are used to design and build complex conversational experiences. These tasks consist of interconnected nodes that define the flow of the conversation. The nodes include entity nodes to capture user information, message nodes to display messages, script nodes to add JavaScript, and service nodes to make API calls.

Using the Task Execution Framework

The task execution framework enables developers to define multiple tasks representing each use case identified for the virtual assistant. These tasks can also include frequently asked questions using the knowledge graph. The knowledge graph allows the virtual assistant to respond to specific questions with predefined answers. Developers can train the knowledge graph to accurately identify and respond to user questions. Additionally, small Talks can be built within the virtual assistant to handle casual conversations that are not meant for dialogue tasks or knowledge tasks.

Digital Views and UI Forms

Digital views provide an appealing and intuitive interface for users, combining conversational and digital experiences. Panels and widgets such as pie charts, tables, and lists can be used to Interact with users. The UI form designer allows developers to Create forms for capturing user details. Components like drop-downs, labels, and separators can be added to create customized forms specific to business needs. The form created using UI forms can also be displayed within panels, enhancing user engagement.

Training the Virtual Assistant

The core AI platform utilizes proprietary NLP technology to train the virtual assistant in understanding user inputs. The machine learning engine, fundamental meaning engine, and knowledge graph engine work together to accurately detect user intent and extract entities. Multiple training utterances can be provided to the machine learning engine to improve its performance. Synonyms and Patterns can be defined for the fundamental meaning engine to identify intents accurately. The ontology-based knowledge graph engine uses a hierarchical structure to identify key domain terms. The virtual assistant can also be trained to handle interruptions, amend entities, and handle multiple intents at a time. Sentiment analysis can be performed to understand the mood of the user and respond accordingly.

Testing the Virtual Assistant

The virtual assistant can be tested using both utterance testing and batch testing. Utterance testing allows developers to test the virtual assistant on various user inputs, checking how well it matches intents. Developers can upload a test suite for batch testing, which tests multiple utterances against the virtual assistant. Reports are generated to Show the matched intents and failed tasks for each test. This helps developers assess the performance of the virtual assistant and make improvements if needed.

Publishing the Virtual Assistant

Once the virtual assistant is tested and fine-tuned, it can be published to end users. The core AI platform offers integrations with more than 30 channels, including popular messaging, voice, and customer service channels. Channels are enabled to allow users to connect with the virtual assistant. Developers can also verify and configure various bot settings before publishing, ensuring a seamless user experience.

Configuring the Virtual Assistant

The virtual assistant can be further configured using the config settings and bot management options. Config settings allow developers to define general settings, PII redaction, authorization profiles, and variables. The bot management section allows developers to share the virtual assistant with other developers, export or import a bot, and track different versions of the virtual assistant. This section also includes change logs and settings for managing the virtual assistant.

Analyzing Performance and User Interactions

Monitoring the performance and user interactions with the virtual assistant is crucial for providing a high-quality service. The core AI platform offers conversation flows and custom dashboards for analyzing performance. Conversation flows provide a visual representation of the user Journey, allowing developers to analyze user interactions with the bot. The metrics report provides detailed insights into matched intents, failed tasks, and API performance. Custom dashboards can be created to generate specific reports based on business needs. These reports help organizations understand bot performance, identify areas for improvement, and make data-driven decisions.

Conclusion

The core AI platform offers a robust and end-to-end solution for building virtual assistants. From designing conversations to training and testing, the platform provides all the necessary tools and capabilities. With the ability to handle complex use cases, integrate with various channels, and analyze performance, the core AI platform empowers organizations to deliver cutting-edge virtual assistants that exceed customer expectations. By leveraging the low code no code platform, enterprises can accelerate their AI initiatives and provide seamless conversational experiences to users. Explore the core AI platform and unlock the full potential of virtual assistants in your organization.

Highlights

• Core AI Platform offers a comprehensive solution for building virtual assistants
• Design conversations using the storyboard feature for natural and complex interactions
• Build virtual assistants using the task execution framework and diverse tasks
• Enhance user experience with digital views, UI forms, and interactive components
• Train virtual assistants using machine learning, fundamental meaning, and knowledge graph models
• Test and analyze performance using utterance and batch testing, conversation flows, and metrics
• Publish virtual assistants to popular messaging and voice channels
• Configure virtual assistants with various settings and manage them through the bot management console
• Analyze the performance and user interactions using conversation flows and custom dashboards
• Core AI platform provides a low code no code solution for rapid development and deployment of virtual assistants

FAQ

Q: Can the virtual assistant understand different languages? A: Yes, the Core AI platform supports more than 19 languages, and the virtual assistant can auto-detect and switch languages based on user utterances.

Q: How can I design a complex conversational experience? A: You can use the task execution framework to create dialogue tasks with interconnected nodes, allowing you to design multi-turn conversations.

Q: Can the virtual assistant handle interruptions in the middle of a conversation? A: Yes, the interruption handling feature enables the virtual assistant to identify new intents even if the user interrupts the current conversation.

Q: Can I analyze the sentiment of user utterances? A: Yes, the sentiment analysis feature allows the virtual assistant to understand the mood of the user and respond accordingly.

Q: How can I test the virtual assistant before publishing? A: The Core AI platform provides utterance testing and batch testing options to test the virtual assistant's performance on various user inputs.

Q: Can I customize and configure the virtual assistant's behavior? A: Yes, the config settings and bot management options allow you to configure general settings, define authorization profiles, and manage the virtual assistant's behavior.

Q: How can I analyze the performance of the virtual assistant? A: The conversation flows and metrics reports provide insights into user journeys, matched intents, failed tasks, and API performance, enabling organizations to monitor and optimize bot performance.