Project Phoenix is a web-based platform designed to integrate natural disaster data with cyber threat intelligence to deliver trusted, real-time risk insights.
It addresses a critical problem: during disasters, cyber threats (misinformation, phishing, scams) increase alongside physical risks.
Phoenix provides a unified system for:
- Detecting risks
- Scoring threats
- Delivering verified alerts
-
Platform: Web Application
-
Primary Users:
- General Public (awareness & alerts)
- Disaster Management Authorities (decision support)
- Build structured hazard + cyber datasets for PHOENIX risk modelling
- Maintain the AI003 cleaning pipeline for validation, duplicate handling, missing value checks, and cleaned CSV/report generation
- Maintain the AI004 feature engineering pipeline for hazard, cyber, temporal, geo, risk, and anomaly features
- Generate synthetic hazard-cyber records for model testing and controlled scenario validation
- Use the reusable AI008 training pipeline for sklearn and PyTorch experiments, checkpoints, reports, and TensorBoard logging
- Develop and compare anomaly detection models including Autoencoder, Isolation Forest, LOF, and One-Class SVM
- Maintain the core PHOENIX risk/correlation models, including Gradient Boosting and XGBoost experiments
- Support AI013 time-series forecasting for wildfire/fire-risk prediction using daily aggregated fire and weather features
- Prepare backend-ready inference outputs and sample JSON records for risk assessment integration
- Identify and integrate reliable data sources (BE001)
- Build ingestion and storage pipelines (BE002, BE005)
- Design scalable database architecture (BE003)
- Develop APIs to serve processed risk data (BE007)
- Enable dashboard integration and data aggregation (BE008)
- Implement monitoring, logging, and deployment systems (BE009–BE012)
- Build notification systems for alerts (BE013)
- Threat Data Finalisation: Prepare a clean, validated, and standardised dataset of disaster-relevant threats (CY001)
- System Architecture Understanding: Analyse and document the system architecture, components, and key assets (CY002)
- Data Flow Diagram: Create a data flow diagram showing how data moves, including entry points and trust boundaries (CY003)
- Threat Identification: Identify potential attackers, vulnerabilities, and attack scenarios affecting the system (CY005)
- STRIDE Threat Modelling: Apply the STRIDE model to assess threats, risks, and corresponding mitigations (CY006)
- TEAVS API Design: Design and document the API structure, endpoints, and alert handling processes (CY007)
- Authentication & Authorisation: Define and design secure authentication and role-based access control mechanisms (CY008)
- Input Validation & Security Rules: Establish rules to validate inputs and protect against common security vulnerabilities (CY009)
- Rate Limiting & Abuse Prevention: Implement controls to limit request rates and prevent abuse or malicious activity (CY010)
- Cryptographic Design: Design secure cryptographic processes, including hashing, signatures, and key management (CY011)
- Security Architecture Design: Develop a comprehensive security architecture integrating all security components (CY012)
- API Development: Build and test the API with full functionality and database integration (CY013)
- Authentication Implementation: Implement and test secure authentication and authorisation mechanisms (CY014)
- Cryptography Implementation: Implement and verify cryptographic features to ensure data integrity and security (CY015)
- Endpoint Security Testing: Test API endpoints for vulnerabilities and document security findings (CY016)
- Logging & Monitoring: Implement logging and monitoring to detect and respond to suspicious activity (CY017)
- Incident Response Workflow: Define a structured process for detecting, managing, and communicating incidents (CY018)
- AI/ML Integration: Integrate and secure AI/ML components within the system while validating outputs (CY019)
- Risk Scoring System: Develop and validate a system to assess and score risks based on threats (CY020)
- Governance & Compliance: Ensure the system meets privacy, ethical, and regulatory compliance requirements (CY021)
- Cyber Resilience Blueprint: Produce a comprehensive blueprint outlining system design, security, and best practices (CY022)
-
Define user journeys for both public and authorities (FE001)
-
Design intuitive low-fidelity and high-fidelity interfaces (FE002–FE006)
-
Build a responsive dashboard for:
- Risk visualisation
- Alerts
- Threat insights
-
Ensure accessibility and usability for real-world emergency scenarios
Data Sources → Data Cleaning → Feature Engineering → AI/ML Models → Risk Engine → API → Web Dashboard
↓
Verified Alert System
- 🌍 Hazard + Cyber Data Integration
⚠️ Real-Time Risk Detection- 📊 Risk Scoring & Forecasting
- 🔐 Verified & Secure Alerts
- 📡 API-driven architecture
- 🖥️ Web-based dashboard
- Languages: Python, JavaScript
- AI/ML: scikit-learn, PyTorch, XGBoost
- Data Processing: Pandas, NumPy, YAML/JSON configs
- Model Types: Random Forest, Gradient Boosting, XGBoost, Isolation Forest, LOF, One-Class SVM, Autoencoder, LSTM forecasting
- Experimentation: Jupyter Notebook, TensorBoard, checkpointing, model reports
- Backend: FastAPI / Flask
- Frontend: Web (React / similar)
- Database: PostgreSQL / MongoDB
- Cloud: AWS
- Messaging: MQTT
- Version Control: Git, GitHub, branch-based PR workflow
🚧 Active Development / Integration Phase
- AI/ML cleaning, feature engineering, synthetic data, training pipeline, anomaly detection, core model, and forecasting workstreams are in progress
- Reusable model training pipeline supports sklearn and PyTorch workflows
- Backend-ready sample records and inference structures are being prepared for integration
- Sprint-based execution continues across AI/ML, backend, cybersecurity, and frontend teams
To build a platform that bridges disaster intelligence and cyber security, enabling:
- Faster and more reliable emergency response
- Reduced impact of misinformation and cyber threats
- Data-driven decision making for authorities and the public