How kesavaai/modus-ai-intelligence-graph works
This is a workforce-impact intelligence application: users explore process-to-activity-to-role-to-skill relationships, inspect automation and payroll exposure, and review reskilling pathways; they can also submit business text for extraction and graph persistence. The frontend includes local JSON-backed serverless routes alongside a FastAPI/Neo4j backend, so these are shown as separate paths rather than assumed to be wired together. Backend pipeline details are based on sampled excerpts; other capabilities are included where documented.
Main entry points: page.tsx Dashboard · routes.py Backend API routes · main.py FastAPI application · neo4j_client.py Graph client
Components 19
User experience
- Dashboard
frontend/src/app/page.tsx - Graph canvas
frontend/src/components/GraphCanvas.tsx - BI analytics
frontend/src/components/BiAnalyticsDashboard.tsx - Ingestion modal
frontend/src/components/SurpriseRecordModal.tsx - Impact drawer
frontend/src/components/CascadeImpactDrawer.tsx
Frontend routes
- Graph route
frontend/src/app/api/v1/graph/all/route.ts - Bundled graph data
frontend/src/data/graph_db.json - Cascade route
frontend/src/app/api/v1/intelligence/cascade/route.ts - Ingest route
frontend/src/app/api/v1/process/ingest/route.ts - Health route
frontend/src/app/api/v1/health/route.ts
Backend services
- FastAPI application
backend/app/main.py - Backend API routes
backend/app/api/routes.py - API schemas
backend/app/models/schemas.py
Graph and extraction
- Extraction pipeline
backend/app/workflows/extraction_pipeline.py - Graph client
backend/app/core/neo4j_client.py - Neo4j graph store
- Seed processes
backend/seed_data.json
Other components
- Enterprise user
- Groq language model
Connections 23
- Enterprise user → Dashboard: uses
- Dashboard → Graph route: requests graph
- Graph route → Bundled graph data: reads
- Graph route → Dashboard: returns graph
- Dashboard → Graph canvas: renders graph
- Dashboard → BI analytics: shows analytics
- Dashboard → Impact drawer: opens impact
- Impact drawer → Cascade route: requests impact
- Cascade route → Bundled graph data: reads
- Enterprise user → Ingestion modal: submits text
- Ingestion modal → Ingest route: posts process
- Dashboard → Health route: checks status
- Enterprise user → FastAPI application: API requests
- FastAPI application → Backend API routes: dispatches
- Backend API routes → API schemas: validates
- Backend API routes → Extraction pipeline: runs extraction
- Extraction pipeline → Groq language model: extracts entities
- Extraction pipeline → Graph client: persists result
- Backend API routes → Graph client: queries graph
- FastAPI application → Graph client: connects and closes
- FastAPI application → Seed processes: loads on startup
- Backend API routes → Seed processes: loads on seed request
- Graph client → Neo4j graph store: reads and writes