GeoSentinel AI: An Agentic Digital Twin for Secure Healthcare Infrastructure

Inspiration

Healthcare systems today are becoming increasingly connected through cloud platforms, IoT medical devices, AI models, and digital health records. While these technologies improve patient care, they also introduce new cybersecurity risks, infrastructure failures, and operational challenges.

My inspiration came from two areas:

My previous research in AI, GIS, and public health analytics.
The growing number of cyberattacks on healthcare organizations that directly impact patient safety.

I wanted to build an intelligent system that doesn't just monitor infrastructure but actively understands, predicts, and responds using autonomous AI agents.

What it does

GeoSentinel AI creates a real-time Digital Twin of healthcare infrastructure.

The platform continuously collects data from infrastructure logs, IoT devices, cybersecurity events, environmental sensors, and geospatial datasets.

Multiple AI agents collaborate to:

Detect cyber threats
Predict equipment failures
Monitor environmental health risks
Assess hospital operational status
Generate recommendations for administrators
Visualize everything on an interactive GIS dashboard

Rather than simply displaying alerts, the agents reason about system health and recommend corrective actions.

How we built it

The platform consists of several modular components.

AI Layer

Agentic AI architecture
LLM-powered reasoning
Multi-agent orchestration

Backend

FastAPI
Python
PostgreSQL/PostGIS
REST APIs

Frontend

React
TypeScript
Tailwind CSS
Map visualization dashboard

Geospatial Intelligence

OpenStreetMap
GIS layers
Environmental datasets
Risk heatmaps

Observability

OpenTelemetry
SigNoz
Distributed tracing
Infrastructure monitoring

Challenges we ran into

The biggest challenge was integrating heterogeneous data sources.

Healthcare infrastructure generates logs, sensor readings, cybersecurity alerts, and geospatial information in different formats. Designing AI agents capable of understanding these sources while maintaining a shared understanding of system state required careful orchestration.

Another challenge was balancing autonomous decision-making with explainability. Every recommendation generated by the agents must be transparent and supported by evidence.

Accomplishments that we're proud of

Built a modular multi-agent architecture.
Combined Digital Twin technology with GIS.
Integrated observability into AI workflows.
Created a unified dashboard for cyber, infrastructure, and public health monitoring.
Designed a scalable architecture suitable for hospitals and smart cities.

What we learned

This project reinforced that modern healthcare resilience requires collaboration between AI, cybersecurity, cloud infrastructure, and geospatial intelligence.

We also learned that observability is essential for trustworthy AI systems. Without visibility into model decisions and infrastructure health, autonomous systems cannot be safely deployed.