Hypaterra Operational Documentation

Learn how to leverage geospatial mapping, predictive timelines, and automated alert engines to monitor global threats.

Platform Overview

Hypaterra is an advanced geospatial intelligence platform designed for analysts to track, map, and predict global events ranging from geopolitical shifts to epidemiological outbreaks.

The platform consumes data from decentralized intelligence feeds, cleans and standardizes the inputs, and maps them in a unified 3D-globe interface. With built-in analytical dashboards, you can move rapidly from high-level global monitoring to granular statistical analysis.

Intelligence Observatory

The **Intelligence Observatory** is the primary real-time monitoring interface.

Data Rendering

Events are plotted onto an interactive 3D WebGL globe. Markers are sized and colored based on threat severity (Critical = Red/Pulsing, Moderate = Blue). When clustering occurs in high-density zones, the markers adapt automatically.

Filtering & Context

Analysts can use the sidebar to filter the globe specifically by region or event class (e.g. "Geopolitical" vs "Health News"). Selecting a marker opens a dedicated detail card providing immediate context about the incident, original source, and timestamp.

Event Forecaster

The Forecaster shifts the platform from *reactive monitoring* to *proactive prediction* using historical pattern aggregation.

The Algorithm

Instead of merely displaying what has happened, the Forecaster analyzes the trailing dataset (e.g., 90 days of records). It builds a threat-calculus based on three distinct matrices:

  • Density: How geographically concentrated are recent events of a similar syntax?
  • Recency Weighting: A logarithmic scale giving precedence to events that happened in the last 48 hours versus last month.
  • Severity Scale: Critical events (e.g., magnitude 8+ earthquakes) have a larger cascading weight than moderate occurrences.

By overlapping these weights, HypaTerra projects a geographical **Confidence Score (Probability)**.

Timeline Scrubber

At the bottom of the Forecaster map, the Timeline Scrubber allows you to fast-forward into the prediction window. Based on historical spread velocity, hotspots will actively pulse, grow, or decay exactly when the algorithm predicts their peak activity day.

Rule-Based Alerts

Monitoring the globe manually 24/7 is impossible. The Alert Engine allows you to establish autonomous watch-rules.

Configuring Rules

Navigate to your Operative Profile and access "Active Subscriptions" to draft a new alert. A rule can be as generic as "All Critical Events in Asia" or incredibly granular, such as "Only Economic events in Japan containing the keyword 'inflation'".

Notification Loop

When the background ingestion task pulls new intelligence, it parses the metadata against your active rules. A positive match produces an immediate system trigger which is permanently archived and visible inside your Profile dashboard feed.

Decision Engine

The HypaTerra Decision Engine is a powerful backend rule-evaluation system. Unlike the simple Alert Engine (which just triggers notifications), the Decision Engine acts as a dynamic logic gateway that assigns confidence scores, calculates risk levels, and generates actionable intelligence recommendations for specific events.

Rule Structure & Keys

Analysts can create complex logical rules via the Django Admin using the DecisionRule model. The engine parses JSON conditions to evaluate an event. Here is a breakdown of the critical JSON keys:

  • field: The property of the event to evaluate (e.g., event_type, severity, region, source).
  • operator: The comparison logic. Supported operators include:
    • eq: Exact match.
    • in: The event's value is inside a provided list.
    • contains: The event's value contains the substring (case-sensitive).
    • icontains: Case-insensitive substring match.
    • gt / lt: Greater than or Less than (for numerical comparisons).
  • value: The target threshold, string, or list to compare against.
  • actions: An array of outcomes triggered when the rule matches. Specify a risk_level (e.g., "CRITICAL", "HIGH") or a recommendation string to display in the UI.

When an event is analyzed via the Intelligence Observatory, it is parsed through all active Decision Rules, compiling the triggered protocols into a single, comprehensive situational report.

Radio Intercepts

Often, the fastest intelligence vector during an incident is localized unencrypted FM radio signals.

The Radio Observatory maps over 30,000 global FM endpoints. By switching the globe into Radio mode, analysts can click on emitting sources in target regions and open a direct audio socket streaming the live broadcast, which is crucial for on-the-ground context verification.

Infrastructure Intelligence

The Infrastructure Observatory maps the physical backbone of the global digital economy, correlating hardware residency with organizational dependency.

Compute Hubs

We track over 1,000 global data centers, including Tier III and Tier IV facilities. Each node is enriched with provider metadata (Equinix, Digital Realty, AWS) and estimated power capacity.

Subsea Connectivity

The platform renders the global submarine cable network as animated data vectors. Analysts can identify critical landing stations where trans-oceanic fiber meets terrestrial networks—often the highest-risk chokepoints in the global supply chain.

Silicon Guard (Risk Engine)

The Silicon Guard engine is a predictive analytics layer that bridges the gap between software AI models and physical hardware.

Heuristic Inference

Because organizational compute residency is often proprietary, Silicon Guard uses a heuristic engine to infer links based on documented cloud partnerships (e.g., OpenAI/Azure) and geographic proximity to high-capacity hubs.

Connectivity Scoring

The engine calculates a Geospatial Connectivity Score (0-100) for any coordinate, evaluating its resilience based on proximity to redundant subsea landings and diversified data center ownership.

Scenario Sandbox

The Scenario Sandbox allows analysts to model "What-If" events and observe their propagation through the World Graph.

Propagation Logic

Simulations use BFS (Breadth-First Search) algorithms to track how a shock at "Ground Zero" travels through entities via relationships like CONTROLS, SUPPLIES, or LOCATED_IN.

Compute Kill Switch

The most advanced simulation mode is the Compute Kill Switch. When activated, the engine simulates a total systemic failure of physical infrastructure hubs linked to the target entities. This provides a dramatic visualization of how hardware outages cascade into AI model unavailability and regional digital darkness.