At the core of eGain Composer is our hybrid knowledge graph approach: we connect to your existing enterprise graphs while maintaining our own internal semantic knowledge model to ensure accurate, governed, and contextually relevant AI answers.
eGain Composer connects to and leverages multiple types of enterprise graphs to deliver contextual, personalized responses:
Consume journey mapping data to understand customer intent, context, and interaction stage. The AI tailors answers based on where the customer is in their journey, ensuring responses are timely and relevant to their current needs.
Integrate with your structured categorization systems to:
- Ground AI responses in your organizational terminology
- Reduce ambiguity in content interpretation
- Ensure compliance with established content standards
- Maintain consistency across all customer touchpoints
Connect to CRM or Customer Data Platform (CDP) systems to personalize responses based on:
- Customer profile and preferences
- Entitlements and service tier
- Historical interactions and support history
- Account-specific context
Integration with product information management enables the AI to reason across:
- Product specifications and features
- Associated troubleshooting procedures
- Related policies and documentation
- Cross-product relationships and dependencies
In agent-assist scenarios, employee graph integration enables:
- Role-based content delivery
- Skills-based routing and guidance
- Access-level appropriate information
- Region or team-specific protocols
In addition to consuming external graphs, eGain maintains an internal semantic knowledge model that serves as a unified, governed layer across all knowledge sources.
- Aggregates and normalizes content from all connected sources
- Creates a unified semantic index for efficient retrieval
- Maintains source attribution and lineage
- Manages synonyms, acronyms, and industry-specific terminology
- Handles multilingual equivalencies
- Adapts to domain-specific language patterns
- Maps semantic relationships (e.g., problem → guidance → resolution)
- Maintains hierarchical and associative connections
- Enables reasoning across related concepts
All knowledge graph integrations respect:
- Existing access controls and permissions
- Data residency and sovereignty requirements
- Privacy and confidentiality classifications
- Enterprise security policies
The internal knowledge model operates within your security perimeter and does not expose data outside established governance boundaries.
Leveraging your enterprise graphs ensures answers reflect your business logic and data