Skip to content

README.md

Latest commit

 

History

History
167 lines (134 loc) · 4.62 KB

README.md

File metadata and controls

167 lines (134 loc) · 4.62 KB
Raw

Supply Chain Resiliency Network Analysis

A comprehensive supply chain risk management platform with network analysis, tariff risk overlay, and AI-powered insights. Built on Databricks with Multi-Agent Supervisor, Knowledge Assistant, and Genie Spaces integration.

Quick Deploy

python deploy.py dev

Features

9-Page Dashboard

  1. Command Center - Real-time KPIs, risk alerts, global map
  2. Max-Flow Analysis - Edmonds-Karp bottleneck detection
  3. Material-Part Network - Bipartite graph, centrality rankings
  4. Supplier-Material Matrix - Risk heatmap, geographic concentration
  5. Tariff-Material Analysis - Tariff exposure, scenario modeling
  6. Supplier-Product Dependencies - Sankey diagram, product risk
  7. Multi-Tier Supplier Network - Hidden dependency detection
  8. Recommendations - AI-generated action items
  9. Executive Briefing - One-page leadership summary

AI Integration

  • Multi-Agent Supervisor (supply-chain-analysis-mas) - Complex analysis
  • Knowledge Assistant (supplytics-knowledge-assistant) - Document search
  • Genie Spaces - Natural language SQL queries:
    • supplytics_global_supply_chain
    • supplytics_supplier_material
    • supplytics_product_customer
    • supplytics_tariffmaterial_product
    • supplytics_material_product

Data Sources (Unity Catalog)

  • chris_messer.bronze.* - Raw dimensions and facts
  • chris_messer.silver.* - Graph edges and nodes
  • chris_messer.gold.* - NetworkX-derived metrics

Tech Stack

Layer Technology
Frontend React 18, TypeScript, Vite 5
Styling Tailwind CSS, Framer Motion
Charts Recharts, D3.js, Cytoscape.js
Backend FastAPI, NetworkX
Database Databricks Unity Catalog
AI Multi-Agent Supervisor, Genie
Deployment Databricks Apps (Asset Bundle)

Project Structure

Supply_Chain_Resiliency_Network_Analysis/
├── src/
│   ├── frontend/          # React application
│   │   ├── src/
│   │   │   ├── components/  # UI components
│   │   │   ├── pages/       # 9 dashboard pages
│   │   │   ├── hooks/       # Custom React hooks
│   │   │   └── services/    # API clients
│   │   └── package.json
│   │
│   └── backend/           # FastAPI application
│       ├── app.py         # Main application
│       ├── routers/       # API endpoints
│       └── services/      # Databricks integration
│
├── sql/                   # SQL stored procedures
├── databricks.yml         # Asset bundle config
├── build.py              # Build script
└── deploy.py             # Deployment script

Development Setup

Prerequisites

  • Node.js 18+
  • Python 3.9+
  • Databricks CLI configured

Local Development

  1. Start Backend
cd src/backend
pip install -r requirements.txt
uvicorn app:app --reload --port 8000
  1. Start Frontend
cd src/frontend
npm install
npm run dev
  1. Access

Environment Variables

Create .env in src/backend/:

DATABRICKS_SERVER_HOSTNAME=fe-vm-hls-amer.cloud.databricks.com
DATABRICKS_TOKEN=your-token
DATABRICKS_WAREHOUSE_ID=your-warehouse-id
MULTI_AGENT_ENDPOINT=supply-chain-analysis-mas
KNOWLEDGE_ENDPOINT=supplytics-knowledge-assistant

Deployment

One-Command Deploy

# Development
python deploy.py dev

# Staging
python deploy.py staging

# Production
python deploy.py prod

Manual Build

python build.py

API Endpoints

Endpoint Description
GET /api/health Health check
GET /api/auth/status SSO authentication status
GET /api/network/nodes Graph nodes by type
GET /api/network/edges Graph edges by type
POST /api/analysis/max-flow Run max-flow algorithm
POST /api/analysis/what-if Disruption simulation
POST /api/chat/multi-agent Multi-Agent Supervisor
POST /api/genie/{space}/query Genie Space query

Competitive Landscape

Competing with:

  • Resilinc - Agentic AI, multi-tier mapping
  • Interos - $1B valuation, AI-driven mapping
  • Everstream Analytics - Predictive risk analytics
  • Kinaxis - Control tower, RapidResponse
  • o9 Solutions - Digital Brain platform

Our Differentiators

  1. Native Databricks integration
  2. Open architecture (not locked in)
  3. SQL-native graph analytics (recursive CTEs)
  4. Multi-Agent AI with domain experts
  5. Customizable & cost-effective

License

Internal use only.

Contact

Supply Chain Analytics Team