Nova is an intelligent, next-generation logistics and supply chain management platform. It uses autonomous AI Agents and Retrieval-Augmented Generation (RAG) to automate complex workflows, extract structured data from unstructured documents, and detect supply chain anomalies in real-time.
For a deep dive into the architecture, AI Agent design, and database schemas, please refer to our detailed documentation: 👉 View Detailed Documentation
- 🤖 Autonomous AI Agents: Features specialized agents for Document Extraction, Exception Detection, Validation, and Workflow Routing.
- 📊 Dynamic Dashboard: Real-time analytics built with Recharts, pulling from PostgreSQL.
- 🔍 Semantic Search (RAG): Ask questions about your logistics data in natural language, powered by ChromaDB.
- ⚡ High-Performance Backend: Built on FastAPI and LangGraph for robust AI orchestration.
- 🐳 Dockerized Infrastructure: One-click setup for PostgreSQL, Redis, and pgAdmin.
Follow these steps to run the Nova platform locally.
Open a terminal in the project root and run the Docker services:
start_docker.bat
# Or manually: cd infra && docker-compose up -dThis starts PostgreSQL (Port 5432), Redis (Port 6379), and pgAdmin.
Open a new terminal and run the backend script:
start_backend.bat
# Or manually: cd backend && python main.pyThe FastAPI server will start on port 8000.
Open a new terminal and run the frontend script:
start_frontend.bat
# Or manually: cd frontend && npm run devThe Vite development server will start on port 5173.
To populate the database and ChromaDB with sample data, run the data loaders from the backend directory:
cd backend
python data_loaders/load_dataco.py
python data_loaders/load_fraud.py
python data_loaders/load_sroie.pyHere is a visual walkthrough of the Nova platform's core features:
Overview: The central hub for creating and monitoring intelligent supply chain workflows. The system dynamically routes tasks across different AI agents based on priority and context, minimizing manual intervention.
Overview: An in-depth view of an active workflow. Watch as the Workflow Router agent autonomously orchestrates tasks, passes data between the Document Extractor and Exception Detector, and maintains a seamless logistics pipeline.
Overview: A comprehensive, real-time snapshot of your logistics operations. Track essential metrics such as total workflows, pending approvals, active exceptions, and overall system health in one unified view.
Overview: The Document Extractor AI in action. It automatically ingests uploaded invoices and supply chain documents, intelligently parses the raw text, and extracts structured data for downstream processing.
Overview: Streamlined management for transactions requiring human oversight. When the AI detects ambiguous data or high-value shipments, the Workflow Router flags them here for manual review and approval.
Overview: Real-time fraud and anomaly detection. The Exception Detector agent continuously monitors purchasing patterns and delivery risks, immediately quarantining high-risk or irregular transactions.
Overview: The system configuration page. Manage your user profile and monitor the real-time readiness of your AI agents and ChromaDB RAG (Retrieval-Augmented Generation) collections.
Built with ❤️ for modern supply chain management.