CAP Alerts composer, communication and dissemination tool

Open-source disaster risk intelligence platform for US municipalities — powered by FEMA, USGS & NOAA public data

Real-time GitHub exploit PoC monitor that scores, analyzes, and alerts on newly published CVE exploits before they trend. Built to give defenders a head start.

Geopolitical early warning system for the Iran-USA-Israel crisis. Real-time event ingestion, 7 risk indices, 5 escalation scenarios with explainable alerts. FastAPI + Next.js + PostgreSQL + Celery on Railway.

Simulated hospital deterioration dataset: 10,000 admissions, hourly vitals+labs over a 72h cohort, with ML-ready tables incl. “next 12h” early-warning target.

System detects the firearm or any other lethal weapon carried by any person and alerts the security team to act before any bad incident occurs

12h clinical deterioration early-warning baseline: tabular models (LogReg / HistGradientBoosting / optional XGBoost), simple probability ensemble, and cost-based threshold/policy tables exported for dashboards.

Hospital deterioration early-warning EDA (first 72h): cohort KPIs, event timing, risk-score behavior, physiologic trajectories, and a toy alert-policy view.

Vedura is a government-oriented public health awareness and early signal intelligence platform designed to support authorities with population-level insights while providing verified, multilingual health guidance to citizens.

An EEW app for Android and Windows. Currently in development. iOS and Linux are coming!

基于 Open-Meteo 气象 + Sentinel-2 遥感指数的轻量智慧农业告警管线。可强自定义 ROI/时间范围/生长季门禁与病虫害(规则)阈值，一键生成告警、事件合并与可解释报告。

Hybrid network + ML framework for corporate financial-distress prediction using network features and classic financial indicators.

A multi-domain geophysical correlation model linking Schumann Resonance, crustal energy, TEC anomalies, and EM events.

Delivery Confidence Assistant built a data‑driven Delivery Confidence Assistant that scores forecast‑credibility risk and translates complex delivery signals into clear, actionable insights. The solution combines transparent rules, lightweight modelling and Power BI‑ready outputs to show which tasks are likely to slip, why programmes and w...

Which ICU variables warn of sepsis earliest? A temporal analysis of 40,336 patients that reveals what the model ignores and why.

Deep Learning for ICU Clinical Time-Series Anomaly Detection & Early Warning — sepsis prediction, mortality, and alarm fatigue reduction on MIMIC-III

Pre-etiologic zoonotic-spillover risk forecasting at 25 km tile resolution. Methodology, 25-event validation cohort, 8-layer governance framework, MCP server source, regulatory pathway. Aqta Technologies Limited (Dublin).

GPM-GARCH: Endogenous heavy-tail formation as a bifurcation phenomenon. Early-warning system for CLT breakdown in financial markets. Formal proofs + empirical validation on 7 markets (2002–2026).

Early Slip Predictor focused on identifying early indicators of delivery slippage by analysing capacity pressure and task behaviour across work centres. Using simple machine‑learning techniques and clear capacity metrics, the team demonstrated how likely future slip can be predicted early and translated into understandable risk signals.