VirusWarn-SC2 is a simple Variants of concern aLert system script for SARS-CoV-2.

🌍 A digital health platform designed to detect, monitor, and help prevent outbreaks of water-borne diseases in vulnerable rural and tribal communities. This system integrates mobile reporting (clinics, ASHA workers, volunteers), IoT water quality sensors, and AI/ML models to predict outbreaks, generate alerts, and support health departments with

PMU data monitoring tool for real-time detection of oscillations in power systems.

"Emergency Management and Coordination System" application is a complete end-to-end disaster management application.

Climate Disaster Warning System is a deep learning-based project for detecting wildfires, floods, and sea-level rise using satellite and ground data. It leverages ResNet, Vision Transformer (ViT), and GRACE datasets to support early warning systems and climate research.

AI-powered ICU early warning system predicting sepsis, hypotension, and hemodynamic collapse 2-6 hours before onset. 4-engine ML ensemble (LightGBM, XGBoost, GRU-D, TCN) with physics-based safety net and multi-agent LLM clinical debate.

Multi-source public health intelligence system for epidemiological surveillance and early warning

Detecting Abnormal Markets - Early Warning Systems

Transforming raw EHR data into predictive clinical insights. A deep learning approach to proactive healthcare and targeted patient interventions.

An early warning system for electrical cable fire risk based on current measurement, harmonic analysis and thermal modeling. Data analysis and risk estimation algorithms were developed using Python and MATLAB.

Undergraduate Report presented as a partial requirement for obtaining the degree of Bachelor in Computer Science

🎓 AI-powered early warning system detecting struggling university students weeks before crisis — XGBoost (AUC 0.975) + Claude Sonnet multi-agent + SHAP explainability + Streamlit dashboard

Explainable Academic Early Warning System (Local Web App)

Implementation of the Self-Supervised Spatiotemporal GNN (ST-GNN) for detecting financial contagion and systemic risk using BIS banking data (1977–2023).

National early warning system for Aadhaar operations — district-level risk detection using privacy-preserving aggregated analytics

Early Warning System Projects

Real-time ML system predicting cloudburst risk for 20 Himalayan hotspots in Uttarakhand. Random Forest trained on 79K+ rows of NASA POWER data. Validated against Kedarnath 2013 (85.8%), Joshimath 2021 (15.1% correctly low). Bilingual UI (EN/हिं), interactive risk map.

CPU-based glyph compression for GPU telemetry (116,640:1 ratio, 8-12h fracture lead time)

Early warning system developed for Saibai Island

Early academic risk detection using student engagement data (OULAD) with time-aware machine learning features.