Visual interface for exploring, analyzing, and managing your knowledge graph. Access at http://localhost:3000 after deployment.
Interactive visualizations for discovering patterns and relationships.
Force-directed layout showing concepts as nodes and relationships as edges.
What you can do:
- See which concepts naturally cluster together
- Identify hub concepts with many connections
- Drag nodes to rearrange the layout
- Click a concept to focus on its neighborhood
- Filter by relationship type or ontology
- Color-code by grounding strength
Best for: Initial exploration, discovering unexpected connections, understanding relationship density.
Immersive three-dimensional version with spatial depth.
What you can do:
- Rotate, pan, and zoom through your knowledge space
- See clusters that overlap in 2D but separate in 3D
- Present impressive visualizations to stakeholders
Best for: Large graphs (1000+ concepts), presentations, finding higher-order structures.
Radial tree centered on a source document.
What you can do:
- See exactly what concepts were extracted from a document
- Trace how extracted knowledge connects to other concepts
- Validate extraction quality
- Follow citation trails back to sources
Best for: Source verification, understanding extraction results, audit trails.
Project concepts onto a semantic spectrum between two poles.
What you can do:
- Define opposing poles (e.g., "Modern" ↔ "Traditional")
- See where each concept falls on the spectrum
- Discover which concepts balance opposing viewpoints
- Check if position correlates with grounding strength
Best for: Understanding conceptual dimensions, classification without predefined categories, finding outliers.
3D visualization of all concept embeddings using t-SNE or UMAP with automatic DBSCAN cluster detection.
What you can do:
- See the overall shape of your semantic space
- View auto-detected clusters with TF-IDF-derived names
- Toggle cluster visibility to focus on specific regions
- Switch color palettes (Bold, Warm→Cool, Earth) and sort by name, count, or color
- Right-click any concept for details and to examine in force graph
- Plan analysis based on what you see
Best for: Discovering semantic dimensions, identifying topic clusters, validating embeddings, global overview before detailed exploration.
System-wide analysis of relationship types.
What you can do:
- See which relationship types are heavily used
- Find dormant vocabulary (defined but rarely used)
- Monitor vocabulary health as you ingest documents
- Identify consolidation opportunities
Best for: System health monitoring, vocabulary maintenance, understanding relationship patterns.
Query-specific breakdown of relationships.
What you can do:
- Analyze relationship types within a specific neighborhood
- Compare subgraph vocabulary to system-wide distribution
- Understand why certain concepts cluster together
Best for: Deep-diving into specific areas, validating relationship classification.
Functional workspaces for specific tasks.
Visual query builder for complex graph traversals.
What you can do:
- Build queries by dragging and connecting blocks
- See compiled Cypher alongside your visual design
- Save and reuse query templates
- Preview results as you build
Best for: Complex queries without writing code, learning openCypher, building reusable analysis templates.
Drag-and-drop document ingestion.
What you can do:
- Drop files onto ontology zones
- See cost estimates before processing
- Create new ontologies on the fly
- Batch submit multiple documents
Supported formats: Text, Markdown, PDF, DOCX, PNG, JPG, WEBP
Best for: Building knowledge graphs from document sets, controlled ingestion with cost visibility.
Monitor and manage extraction jobs.
What you can do:
- See job status (pending, processing, completed, failed)
- Approve or cancel jobs before processing
- View cost estimates and actual costs
- Clean up old jobs in bulk
Best for: Workflow control, cost management, debugging failed extractions.
Tabular views and export functionality.
What you can do:
- View explorer results as sortable tables
- Export to CSV or JSON
- Track changes between analyses (delta indicators)
- Copy data to clipboard
Best for: Analysis reports, data science workflows, sharing with non-technical stakeholders.
Manual creation and editing of concepts and edges.
What you can do:
- Create concepts without running extraction
- Add relationships between existing concepts
- Fix extraction errors
- Remove duplicates
Best for: Expert curation, correcting LLM mistakes, adding domain knowledge.
User and system management.
What you can do:
- Manage users and roles (admin only)
- Create and revoke OAuth clients
- Monitor system health
- View job queue statistics
Best for: Multi-user deployments, API key management, system monitoring.
- Start at Embedding Landscape — see overall structure
- Switch to 2D Force Graph — drill into neighborhoods
- Use Polarity Explorer — find semantic dimensions
- Export to Data Export — document findings
- Check Job Queue — verify extraction completed
- Open Document Explorer — see what was extracted
- Review in 2D Force Graph — spot-check relationships
- Fix errors in Graph Editor