Back to Admin
Gartner Critical Capabilities for Conversational AI Platforms — August 2025

Lucy Gartner Compliance Audit

Self-assessment against Gartner's Critical Capabilities framework for Conversational AI Platforms (2025 edition), covering 10 capability domains and 90 individual criteria.

99%
Overall Gartner Compliance Score
88 Fully Compliant
2 Partially Compliant
Natural92%
Voice100%
Multilingual100%
Channel100%
Analytics100%
Privacy100%
AI100%
Agentic100%
Lifecycle95%
Personalisation100%

Action Items (2 partial, 0 gaps)

  • [Natural Language Understanding] Custom NLU model training — Training data ingested via ingestAndProcessTrainingData and LucyScenarioTraining page, but no self-serve NLU retraining UI exposed to hotel operators
  • [Lifecycle Management & Administration] Multi-hotel deployment management — MultiHotelDashboard and MultiHotelDeployment pages exist; hotel_id isolation confirmed. White-label provisioning planned.

Detailed Criteria Breakdown

Gartner definition:Ability to understand natural language input, extract entities, handle intent classification, training and customisation of NLU models.
Intent classification & entity extractionCOMPLIANT

advancedIntentClassification, lucyInnerCore, parseComplexVoiceIntent — multi-turn intent graph built per conversation

Contextual dialogue management (multi-turn)COMPLIANT

handleMultiTurnDialogue, buildEnhancedConversationContext, shortTermContextManager — full turn management stack

Ambiguity resolution & clarificationCOMPLIANT

detectConfidenceAndClarify, handleAmbiguousVoiceCommand, detectSpellOutNeeded

Custom NLU model trainingPARTIAL

Training data ingested via ingestAndProcessTrainingData and LucyScenarioTraining page, but no self-serve NLU retraining UI exposed to hotel operators

Sentiment & emotion detectionCOMPLIANT

enhancedEmotionDetection, analyzeGuestSentimentRealTime, emotionalTrajectoryEngine, lucySentimentAdaptiveResponse — full sentiment pipeline

Code-switching (mid-sentence language shift)COMPLIANT

detectCodeSwitching function purpose-built for this; tested in LucyTwoWayTranslationTester

Self-assessment conducted against Gartner Critical Capabilities for Conversational AI Platforms (August 2025). Evidence sourced from Lucy's codebase, backend function registry, entity schema, and agent configurations. Audit date: May 2026.