GETTING STARTED
Kore.ai XO Platform
Virtual Assistants Overview
Natural Language Processing (NLP)
Concepts and Terminology
Quick Start Guide
Accessing the Platform
Navigating the Kore.ai XO Platform
Building a Virtual Assistant
Help & Learning Resources
Release Notes
Current Version
Recent Updates
Previous Versions
Deprecations
Request a Feature
CONCEPTS
Design
Storyboard
Overview
FAQs
Conversation Designer
Overview
Dialog Tasks
Mock Scenes
Dialog Tasks
Overview
Navigate Dialog Tasks
Build Dialog Tasks
Node Types
Overview
Intent Node
Dialog Node
Dynamic Intent Node
GenAI Node
GenAI Prompt
Entity Node
Form Node
Confirmation Node
Message Nodes
Logic Node
Bot Action Node
Service Node
Webhook Node
Script Node
Process Node
Agent Transfer
Node Connections
Node Connections Setup
Sub-Intent Scoping
Entity Types
Entity Rules
User Prompts or Messages
Voice Call Properties
Knowledge AI
Introduction
Knowledge Graph
Introduction
Terminology
Build a Knowledge Graph
Manage FAQs
Knowledge Extraction
Import or Export Knowledge Graph
Prepare Data for Import
Importing Knowledge Graph
Exporting Knowledge Graph
Auto-Generate Knowledge Graph
Knowledge Graph Analysis
Answer from Documents
Alert Tasks
Small Talk
Digital Skills
Overview
Digital Forms
Digital Views
Introduction
Widgets
Panels
Session and Context Variables
Context Object
Intent Discovery
Train
NLP Optimization
ML Engine
Overview
Model Validation
FM Engine
KG Engine
Traits Engine
Ranking and Resolver
Training Validations
NLP Configurations
NLP Guidelines
LLM and Generative AI
Introduction
LLM Integration
Kore.ai XO GPT Module
Prompts & Requests Library
Co-Pilot Features
Dynamic Conversations Features
Intelligence
Introduction
Event Handlers
Contextual Memory
Contextual Intents
Interruption Management
Multi-intent Detection
Amending Entities
Default Conversations
Conversation Driven Dialog Builder
Sentinment Management
Tone Analysis
Default Standard Responses
Ignore Words & Field Memory
Test & Debug
Overview
Talk to Bot
Utterance Testing
Batch Testing
Conversation Testing
Conversation Testing Overview
Create a Test Suite
Test Editor
Test Case Assertion
Test Case Execution Summary
Glossary
Health and Monitoring
NLP Health
Flow Health
Integrations
Actions
Actions Overview
Asana
Configure
Templates
Azure OpenAI
Configure
Templates
BambooHR
Configure
Templates
Bitly
Configure
Templates
Confluence
Configure
Templates
DHL
Configure
Templates
Freshdesk
Configure
Templates
Freshservice
Configure
Templates
Google Maps
Configure
Templates
Here
Configure
Templates
HubSpot
Configure
Templates
JIRA
Configure
Templates
Microsoft Graph
Configure
Templates
Open AI
Configure
Templates
Salesforce
Configure
Templates
ServiceNow
Configure
Templates
Stripe
Configure
Templates
Shopify
Configure
Templates
Twilio
Configure
Templates
Zendesk
Configure
Templates
Agents
Agent Transfer Overview
Custom (BotKit)
Drift
Genesys
Intercom
NiceInContact
NiceInContact(User Hub)
Salesforce
ServiceNow
Configure Tokyo and Lower versions
Configure Utah and Higher versions
Unblu
External NLU Adapters
Overview
Dialogflow Engine
Test and Debug
Deploy
Channels
Publishing
Versioning
Analyze
Introduction
Dashboard Filters
Overview Dashboard
Conversations Dashboard
Users Dashboard
Performance Dashboard
Custom Dashboards
Introduction
Custom Meta Tags
Create Custom Dashboard
Create Custom Dashboard Filters
LLM and Generative AI Logs
NLP Insights
Task Execution Logs
Conversations History
Conversation Flows
Conversation Insights
Feedback Analytics
Usage Metrics
Containment Metrics
Universal Bots
Introduction
Universal Bot Definition
Universal Bot Creation
Training a Universal Bot
Universal Bot Customizations
Enabling Languages
Store
Manage Assistant
Team Collaboration
Plan & Usage
Overview
Usage Plans
Templates
Support Plans
Invoices
Authorization
Conversation Sessions
Multilingual Virtual Assistants
Get Started
Supported Components & Features
Manage Languages
Manage Translation Services
Multiingual Virtual Assistant Behavior
Feedback Survey
Masking PII Details
Variables
Collections
IVR Settings
General Settings
Assistant Management
Manage Namespace
Data
Overview
Data Table
Table Views
App Definitions
Data as Service
HOW TOs
Build a Travel Planning Assistant
Travel Assistant Overview
Create a Travel Virtual Assistant
Design Conversation Skills
Create an ‘Update Booking’ Task
Create a Change Flight Task
Build a Knowledge Graph
Schedule a Smart Alert
Design Digital Skills
Configure Digital Forms
Configure Digital Views
Train the Assistant
Use Traits
Use Patterns
Manage Context Switching
Deploy the Assistant
Use Bot Functions
Use Content Variables
Use Global Variables
Use Web SDK
Build a Banking Assistant
Design Conversation Skills
Create a Sample Banking Assistant
Create a Transfer Funds Task
Create a Update Balance Task
Create a Knowledge Graph
Set Up a Smart Alert
Design Digital Skills
Configure Digital Forms
Configure Digital Views
Add Data to Data Tables
Update Data in Data Tables
Add Data from Digital Forms
Train the Assistant
Composite Entities
Use Traits
Use Patterns for Intents & Entities
Manage Context Switching
Deploy the Assistant
Configure an Agent Transfer
Use Assistant Functions
Use Content Variables
Use Global Variables
Intent Scoping using Group Node
Analyze the Assistant
Create a Custom Dashboard
Use Custom Meta Tags in Filters
Migrate External Bots
Google Dialogflow Bot
APIs & SDKs
API Reference
API Introduction
Rate Limits
API List
koreUtil Libraries
SDK Reference
SDK Introduction
Web SDK
How the Web SDK Works
SDK Security
SDK Registration
Web Socket Connect and RTM
Tutorials
Widget SDK Tutorial
Web SDK Tutorial
BotKit SDK
BotKit SDK Deployment Guide
Installing the BotKit SDK
Using the BotKit SDK
SDK Events
SDK Functions
Tutorials
BotKit - Blue Prism
BotKit - Flight Search Sample VA
BotKit - Agent Transfer
  1. Home
  2. Integrations
  3. Test and Debug

Test and Debug

The following sections describe how you can test and debug your Virtual Assistant to validate the working of External NLU.

Talk to Bot

After completing the steps in Adapter Configuration and enabling the API, to check if the intent identification is done by External NLU adapter or not, you can follow these steps:

    1. Go to your VA and click Talk to Bot.
    2. Enter a user utterance which matches the intent entered in the Dialogflow Essentials page.
    3. In the Debug Log panel of the user conversation, you can see a message that, intent identified using External NLU is initiated.


You can also analyze the Debug Log, NL Analysis, and Session Context & Variables to understand the intent detection mechanism.

Debug Log

If the external NLU is enabled, the Debug Log displays whether the intent is detected by External NLU or not before the ‘Intent node initiated’ debug log entry. In case of call failure, an error response is displayed in the debug log entry.

This is displayed only if an external NLU is enabled. If the API call fails, it is also indicated in the debug log statement.

    • Intent detected by the External NLU engine is always marked as a Definitive Match.
      If intent is not detected by the External NLU engine, it is tracked in Debug Log.
    • The debug logs display the request and response payload for the calls made to the LLM by the Dynamic Conversation features only.
    • Under Session Context & Variables, you can see the variables for intents and entities in detail as shown in the following screenshots.

 

 

NLP Analysis

By using NL analysis, you can see whether the intent was identified or not by an external NLU engine. This works from all the workflows that provide the NLP analysis. For example – Talk to Bot, Utterance Testing, Batch Testing, Health and Monitoring, and NLP Insights.

See the following sections to understand more details.

Utterance Testing

Utterance Testing also provides NLP Analysis. The Matched Intents and Fields/Entities tables that provide an easy way to update training data like synonyms, patterns, etc are not displayed if the Intent is identified on the external NLU engine.

    • If the intent is detected by an external engine, the utterance is not sent to the Kore.ai NLP engine.

    • Click the View Response link to view the API response in the JSON format, received from the external NLU engine in both cases of whether intent is detected or not detected by the external NLU.

    • If the intent is not detected by the external engine, the Kore.ai NLU engine acts as a fallback. So, in this case, the NLP analysis shows the analysis of intent detection results from our three engines along with the ranking and resolver. For more information, see Ranking and Resolver.

    • If the intent is detected by the external engine but cannot be mapped to any intent on the XO platform, even then the Kore.ai NLU engine acts as a fallback. In this case, the NLP analysis displays a relevant message along with the API response containing the detected Intents and parameters.

Batch Testing

The test case result for True positive, False Positive, True Negative, and False Negative(TP, FP, TN, FN) scenarios is generated based on the net outcome of a combined result of External NLU and Kore.ai NLU engine.

For example, irrespective of whether the expected intent is detected by the External NLU or Kore.ai engine, the test result is TP. If the net result is a wrong intent identified, the test result should be FP. In case of no intent detected by any of the engines, it is classified as TN, when no intent was expected but some intent was detected, then it is FN.

Note: There is no common base to compare the scores of all the engines, hence scores sent by the external NLU are not considered.

For any intent detected by the external NLU, it should be considered that the thresholds have been met and in the ‘Matched Intent Score’ column, this is indicated as ExtNLU (without any score against it) in place of ML, FM, or KG scores. This can be viewed in the downloaded CSV under NLU of Batch Testing. To know more, see Batch Testing.

If the ExtNLU returns multiple intents, the thresholds met for all those intents are considered and accordingly the test result is concluded.

Health and Monitoring

In Health and Monitoring go to Test Cases —> Detailed Analysis to see the Intents and Entities identified by External NLU, without any scores. The Traits engine is not dependent on the external NLU.

Note: If the external NLU is enabled, some of the recommendations may not be relevant because they are for the Kore.ai NLU engines. The recommendations are displayed based on the training done on the Kore.ai Platform.


For more information, see Virtual Assistant Health and Monitoring.

NLP Insights

NLP Insights have Intent Identified, Intent Not Identified utterances and their NLP analysis details captured. You can view the details of all the failed External NLU API calls in the debug logs of NLP Insights.

For more information, see NLP Insights.

Conversation Behavior

This list describes the conversation behavior while using external NLU adapter:

  • When an intent is identified by the external engine, the corresponding task on the Kore.ai Platform is triggered based on the name mapping.
  • Quality of the intent, entity, or FAQ detection depends on the training and capabilities of the external engine
  • All the other conversation flow behaviors are handled by the Platform. For e.g.:
    • Re-prompts, error prompts, message responses, widgets, supported channels, channel-specific experience customizations, authorization for service calls, alerts, small talk
    • Language Detection
    • Sentiment Analysis
    • Traits identification
    • Decisions to scan for sub-intent, follow-up intents
    • Interruption handling as per settings
    • Entity over Intent or Intent over Entity as per settings
    • PII Masking/Redaction
    • Execution of all the nodes of a Dialogue Task
    • Context Object construction
    • Response messages and fulfillment
    • Logical flows
    Note: Information like the response messages and fulfillment from the external engine is ignored.
  • When external NLU is enabled, in the interruption handling only one behavior, “Switch to a new task without any notification to the user and discard the current task”, is supported. To know more information, see Interruption Management Settings.
  • If the external NLU API returns an error or the API cannot be reached, a fallback intent gets activated with the standard response as, Error in continuing the conversation due to incorrect bot definition. To know more information, see Standard Responses.
  • Note: The Intent and entity mapping work based on a case insensitive comparison of the names defined on the two platforms.
  • For the entity identification to work properly, it is recommended to define the entity nodes in the same sequence on the XO Platform and the external engine.
  • If the external NLU fails to detect the entity, then as a fallback, an attempt is made to detect the same on the Kore.ai NLP engine.

Follow-up and Sub-Intents Identification

  • If all the required parameters are available in the context, the Follow-up intent scope gets activated on the External NLU. Once the follow-up intent scope gets active on the Platform, the utterance is sent to External NLU for intent identification.
  • Similarly, when the scope for sub-intent is active on the Kore.ai Platform, the utterances should be sent to the external NLU for intent detection along with the parent intent context.

Test and Debug

The following sections describe how you can test and debug your Virtual Assistant to validate the working of External NLU.

Talk to Bot

After completing the steps in Adapter Configuration and enabling the API, to check if the intent identification is done by External NLU adapter or not, you can follow these steps:

    1. Go to your VA and click Talk to Bot.
    2. Enter a user utterance which matches the intent entered in the Dialogflow Essentials page.
    3. In the Debug Log panel of the user conversation, you can see a message that, intent identified using External NLU is initiated.


You can also analyze the Debug Log, NL Analysis, and Session Context & Variables to understand the intent detection mechanism.

Debug Log

If the external NLU is enabled, the Debug Log displays whether the intent is detected by External NLU or not before the ‘Intent node initiated’ debug log entry. In case of call failure, an error response is displayed in the debug log entry.

This is displayed only if an external NLU is enabled. If the API call fails, it is also indicated in the debug log statement.

    • Intent detected by the External NLU engine is always marked as a Definitive Match.
      If intent is not detected by the External NLU engine, it is tracked in Debug Log.
    • The debug logs display the request and response payload for the calls made to the LLM by the Dynamic Conversation features only.
    • Under Session Context & Variables, you can see the variables for intents and entities in detail as shown in the following screenshots.

 

 

NLP Analysis

By using NL analysis, you can see whether the intent was identified or not by an external NLU engine. This works from all the workflows that provide the NLP analysis. For example – Talk to Bot, Utterance Testing, Batch Testing, Health and Monitoring, and NLP Insights.

See the following sections to understand more details.

Utterance Testing

Utterance Testing also provides NLP Analysis. The Matched Intents and Fields/Entities tables that provide an easy way to update training data like synonyms, patterns, etc are not displayed if the Intent is identified on the external NLU engine.

    • If the intent is detected by an external engine, the utterance is not sent to the Kore.ai NLP engine.

    • Click the View Response link to view the API response in the JSON format, received from the external NLU engine in both cases of whether intent is detected or not detected by the external NLU.

    • If the intent is not detected by the external engine, the Kore.ai NLU engine acts as a fallback. So, in this case, the NLP analysis shows the analysis of intent detection results from our three engines along with the ranking and resolver. For more information, see Ranking and Resolver.

    • If the intent is detected by the external engine but cannot be mapped to any intent on the XO platform, even then the Kore.ai NLU engine acts as a fallback. In this case, the NLP analysis displays a relevant message along with the API response containing the detected Intents and parameters.

Batch Testing

The test case result for True positive, False Positive, True Negative, and False Negative(TP, FP, TN, FN) scenarios is generated based on the net outcome of a combined result of External NLU and Kore.ai NLU engine.

For example, irrespective of whether the expected intent is detected by the External NLU or Kore.ai engine, the test result is TP. If the net result is a wrong intent identified, the test result should be FP. In case of no intent detected by any of the engines, it is classified as TN, when no intent was expected but some intent was detected, then it is FN.

Note: There is no common base to compare the scores of all the engines, hence scores sent by the external NLU are not considered.

For any intent detected by the external NLU, it should be considered that the thresholds have been met and in the ‘Matched Intent Score’ column, this is indicated as ExtNLU (without any score against it) in place of ML, FM, or KG scores. This can be viewed in the downloaded CSV under NLU of Batch Testing. To know more, see Batch Testing.

If the ExtNLU returns multiple intents, the thresholds met for all those intents are considered and accordingly the test result is concluded.

Health and Monitoring

In Health and Monitoring go to Test Cases —> Detailed Analysis to see the Intents and Entities identified by External NLU, without any scores. The Traits engine is not dependent on the external NLU.

Note: If the external NLU is enabled, some of the recommendations may not be relevant because they are for the Kore.ai NLU engines. The recommendations are displayed based on the training done on the Kore.ai Platform.


For more information, see Virtual Assistant Health and Monitoring.

NLP Insights

NLP Insights have Intent Identified, Intent Not Identified utterances and their NLP analysis details captured. You can view the details of all the failed External NLU API calls in the debug logs of NLP Insights.

For more information, see NLP Insights.

Conversation Behavior

This list describes the conversation behavior while using external NLU adapter:

  • When an intent is identified by the external engine, the corresponding task on the Kore.ai Platform is triggered based on the name mapping.
  • Quality of the intent, entity, or FAQ detection depends on the training and capabilities of the external engine
  • All the other conversation flow behaviors are handled by the Platform. For e.g.:
    • Re-prompts, error prompts, message responses, widgets, supported channels, channel-specific experience customizations, authorization for service calls, alerts, small talk
    • Language Detection
    • Sentiment Analysis
    • Traits identification
    • Decisions to scan for sub-intent, follow-up intents
    • Interruption handling as per settings
    • Entity over Intent or Intent over Entity as per settings
    • PII Masking/Redaction
    • Execution of all the nodes of a Dialogue Task
    • Context Object construction
    • Response messages and fulfillment
    • Logical flows
    Note: Information like the response messages and fulfillment from the external engine is ignored.
  • When external NLU is enabled, in the interruption handling only one behavior, “Switch to a new task without any notification to the user and discard the current task”, is supported. To know more information, see Interruption Management Settings.
  • If the external NLU API returns an error or the API cannot be reached, a fallback intent gets activated with the standard response as, Error in continuing the conversation due to incorrect bot definition. To know more information, see Standard Responses.
  • Note: The Intent and entity mapping work based on a case insensitive comparison of the names defined on the two platforms.
  • For the entity identification to work properly, it is recommended to define the entity nodes in the same sequence on the XO Platform and the external engine.
  • If the external NLU fails to detect the entity, then as a fallback, an attempt is made to detect the same on the Kore.ai NLP engine.

Follow-up and Sub-Intents Identification

  • If all the required parameters are available in the context, the Follow-up intent scope gets activated on the External NLU. Once the follow-up intent scope gets active on the Platform, the utterance is sent to External NLU for intent identification.
  • Similarly, when the scope for sub-intent is active on the Kore.ai Platform, the utterances should be sent to the external NLU for intent detection along with the parent intent context.
Menu