How semantica-agi/semantica works
Semantica is graph-native infrastructure for turning enterprise data into governed context and knowledge graphs, then querying them for reasoning, decisions, and auditable provenance. The documented user workflow is ingest, extract and model data, build/query graphs, and inspect results through the Knowledge Explorer; optional ontology, memory, and agent-tool capabilities extend that core. Relationships are cited to the README where it documents product-level flows, or to indexed source references where they show module connections; unsampled route behavior is not inferred beyond those references.
Main entry points: app.py Explorer API · knowledge_graph.py Knowledge graph · context_graph.py Context graph · markdown_resources.py Markdown resources
Components 20
Data acquisition
- Data ingestion
semantica/ingest - Parse and extract
semantica/semantic_extract - Pipeline execution
semantica/pipeline/execution_engine.py
Knowledge modeling
- Ontology engine
semantica/ontology/engine.py - Knowledge graph
semantica/kg/knowledge_graph.py - Graph storage
semantica/graph_store - RDF storage and query
semantica/triplet_store/query_engine.py
Context and reasoning
- Context graph
semantica/context/context_graph.py - Agent memory
semantica/context/agent_memory.py - Deterministic reasoning
semantica/reasoning/reasoner.py - Decision context
semantica/context/decision_context.py
Trust and governance
- Provenance tracking
semantica/provenance/manager.py - Provenance storage
semantica/provenance/storage.py
Knowledge Explorer
- Explorer workspaces
explorer/src/App.tsx - Explorer API
semantica/explorer/app.py - Graph API routes
semantica/explorer/routes/graph.py - Ontology API routes
semantica/explorer/routes/ontology.py - Enrichment routes
semantica/explorer/routes/enrich.py - Markdown resources
semantica/explorer/markdown_resources.py
Other components
- Data / AI engineer
Connections 20
- Data / AI engineer → Data ingestion: provides data
README.md - Data ingestion → Parse and extract: feeds extraction
README.md - Parse and extract → Knowledge graph: builds graph
README.md - Pipeline execution → Data ingestion: orchestrates workflows
- Ontology engine → RDF storage and query: queries and validates
semantica/ontology/engine.py - Knowledge graph → Graph storage: uses graph storage
- Knowledge graph → RDF storage and query: exports to RDF
- Context graph → Knowledge graph: integrates graph analytics
semantica/context/context_graph.py - Deterministic reasoning → Context graph: reasons over context
- Decision context → Context graph: records decisions
- Decision context → Provenance tracking: tracks decision history
- Provenance tracking → Provenance storage: persists records
semantica/provenance/manager.py - Data / AI engineer → Explorer workspaces: explores results
README.md - Explorer workspaces → Explorer API: uses API
- Explorer API → Graph API routes: registers routes
semantica/explorer/app.py - Explorer API → Ontology API routes: registers routes
semantica/explorer/app.py - Explorer API → Enrichment routes: registers routes
semantica/explorer/app.py - Explorer API → Markdown resources: initializes resources
semantica/explorer/app.py - Markdown resources → Context graph: adapts Markdown access
semantica/explorer/markdown_resources.py - Markdown resources → Agent memory: adapts Markdown access
semantica/explorer/markdown_resources.py