Workspace Validation

Introduction: Workspace validation feature will validate user actions in a workspace against a set of pre defined conditions to show Warnings & Recommendations. Complying with these pre defined conditions improves overall bot performance for AI powered journeys. Validation features assist users in settings up the workspace according to best practices.

Key Highlights:

Validation icon introduced at workspace level to show warnings & recommendations in separate tabs
Validation conditions are checked after every save & train operation
User is redirected to relevant page to take corrective action on warnings & recommendations
Validation icon is be visible inside the workspace if at least 1 warning or recommendation is there
Error icon is introduced workspace having warning or recommendation on workspace listing page

Validation Conditions:

`Warnings`	`Recommendations`
Condition: No Content Uploaded in content section Title: Content Missing in Content section Description: Add training data in workspace for answering questions	Condition: less than 10 utterances ( AI + Manual) trained for a single Intent - * Title: Less than 10 utterances mapped to intent Description: Sample utterances help improve consistency in intent detection
Condition: No Workspace Instructions for answers in the Settings Tab Title: Instructions for answers missing in Settings section Description: Workspace instructions provides more context to the bot on its role and improves overall behavior	Condition: No Sample Values: #Entity 1, #Entity2 Title: Sample values missing for Entity Description: Improves consistency for certain entity values
Condition: No Entity mapped to Intent * Title: Intent-Entity Mapping Missing Description: Entity detection will be skipped for intent related utterances	Condition: No Extended Values:#Entity 1, #Entity2 Title: Add extended values for Entity Description: Improves handling of input variability and better generalization in recognizing entity
Condition: Precision Threshold is lower than 30% Title: Precision Threshold less than 30% Description: Low threshold may lead to inaccurate answers from bot	Condition: No Manual Utterances mapped to Intent Title: Manual Utterance not added for intent Description: Manual utterances improves accuracy of model prediction