How kaidenhuang/openpalantir works
OpenPalantir ingests documents and relational databases, extracts and enriches entities and relationships, persists a Neo4j knowledge graph, and exposes graph analysis and an AI decision assistant through a React UI and FastAPI backend. The graph preserves the documented ingestion, CDC, task, analysis, and decision branches. Several backend implementations were not sampled, so relationships supported only by product documentation are cited to the README; uncertain wiring is omitted.
Main entry points: main.py REST API · App.tsx Application shell · graph_manager.py Graph manager · database.py Database routes
Components 33
User interface
- Application shell
frontend/src/App.tsx - Database UI
frontend/src/components/DatabaseManagement.tsx - Graph visualization
frontend/src/components/GraphVisualization.tsx - Analysis dashboard
frontend/src/components/AnalysisDashboard.tsx - Decision assistant
frontend/src/components/DecisionAssistant.tsx - Document viewer
frontend/src/components/DocumentViewer.tsx
API and tasks
- REST API
backend/main.py - Database routes
backend/api/routes/database.py - Analysis routes
backend/api/routes/analysis.py - Decision routes
backend/api/routes/decision.py - Task manager
backend/task_management/task_manager.py - Task handlers
backend/task_management/handlers - SQLite metadata
backend/config/database.py
Data ingestion
- Source management
backend/source_management/source_service.py - Document processing
backend/document_processing/document_processor.py - Document indexing
backend/pageindex/page_index.py - Database connections
backend/database_management/database_manager.py - Schema persistence
backend/database_management/database_service.py - Entity extraction
backend/entity_extraction/llm_entity_enhancer.py - CDC synchronization
backend/cdc/cdc_manager.py
Knowledge graph
- Graph manager
backend/knowledge_graph/graph_manager.py - Entity repository
backend/knowledge_graph/entity_repository.py - Relationship repository
backend/knowledge_graph/relationship_repository.py - Neo4j
Analysis and decisions
- Graph analyzer
backend/analysis_engine/analyzer.py - Decision engine
backend/decision_engine/agentic/engine.py - Decision tools
backend/decision_engine/agentic/tools.py - Model client
backend/model_management/model_client.py
Other components
- User
- LLM provider
- Source databases
- Debezium
- Redis Streams
Connections 40
- User → Application shell: uses
README.md - Application shell → Database UI: renders
frontend/src/App.tsx - Application shell → Graph visualization: renders
frontend/src/App.tsx - Application shell → Analysis dashboard: renders
frontend/src/App.tsx - Application shell → Decision assistant: renders
frontend/src/App.tsx - Application shell → Document viewer: renders
frontend/src/App.tsx - Database UI → REST API: requests
README.md - Graph visualization → REST API: requests
README.md - Analysis dashboard → REST API: requests
README.md - Decision assistant → REST API: requests
README.md - Document viewer → REST API: requests
README.md - REST API → Database routes: registers
backend/main.py - REST API → Analysis routes: registers
backend/main.py - REST API → Decision routes: registers
backend/main.py - Database routes → Database connections: calls
backend/api/routes/database.py - Database routes → Schema persistence: calls
backend/api/routes/database.py - REST API → Task manager: exposes
README.md - Task manager → Task handlers: dispatches
backend/task_management/task_manager.py - Source management → Document processing: processes documents
- Document processing → Document indexing: indexes documents
- Source management → Entity extraction: enriches entities
backend/source_management/source_service.py - Database connections → Source databases: connects
README.md - Database connections → Schema persistence: extracts schema
backend/database_management/database_manager.py - CDC synchronization → Debezium: captures changes
README.md - Debezium → Source databases: reads changes
README.md - Debezium → Redis Streams: streams events
README.md - CDC synchronization → Redis Streams: consumes events
README.md - Entity extraction → Graph manager: adds graph data
- Graph manager → Entity repository: delegates entities
backend/knowledge_graph/graph_manager.py - Graph manager → Relationship repository: delegates relationships
backend/knowledge_graph/graph_manager.py - Entity repository → Neo4j: reads and writes
backend/knowledge_graph/entity_repository.py - Relationship repository → Neo4j: reads and writes
backend/knowledge_graph/relationship_repository.py - Analysis routes → Graph analyzer: runs analysis
- Graph analyzer → Graph manager: queries graph
- Decision routes → Decision engine: runs assistant
- Decision engine → Decision tools: executes tools
backend/decision_engine/agentic/engine.py - Decision engine → Model client: requests completions
backend/decision_engine/agentic/engine.py - Decision tools → Graph manager: retrieves graph data
- Model client → LLM provider: calls model
backend/model_management/model_client.py - REST API → SQLite metadata: initializes
backend/main.py