Home / Documentation / How It Works

How It Works

A plain-language guide to understanding The Architect platform. No technical jargon required.

This page explains how data flows through the system, how scores are computed, and how every decision is made transparent and auditable.

The Platform Workflow

📡
Step 1

Data Collection

The platform gathers information from trusted public sources — like the World Bank, official company registries, and geographic databases. We never invent or guess data. If information is missing, we show 'Pending' rather than making something up.

  • World Bank provides economic data (GDP, population, education, trade) for 200+ countries
  • GeoNames provides cities, councils, and geographic boundaries
  • Companies House (UK) and SEC EDGAR (US) provide official company information
  • PrismAPI enriches with real-time inflation, interest rates, and trade flows
Automated — refreshed every 12 hours
🧮
Step 2

Score Computation

Each country gets a score from 0 to 100 based on a transparent formula. The formula looks at four areas of national performance, each weighted by importance. The scoring is fully auditable — you can see exactly which data point contributed to each score.

  • Economic Velocity (30%) — How strong is the economy? GDP per capita, jobs, trade
  • Societal Wellness (30%) — How well do people live? Health, education, internet access
  • Geopolitical Stability (25%) — How peaceful? Military spending, conflicts, inequality
  • Resource Autonomy (15%) — How sustainable? Renewable energy, food production, emissions
Computed on-demand with 12h cache
🛡️
Step 3

AI Security Layer

Every interaction with the AI assistant passes through Veea Lobster Trap — a security engine that checks for threats in real time. It protects against prompt injection, data theft, credential leaks, and compliance violations. Think of it as a firewall for AI conversations.

  • 18 security rules across 4 compliance packs (General, HIPAA, SOC2, Finance)
  • Sub-millisecond evaluation — users don't notice any delay
  • Intent detection catches when requests don't match their stated purpose
  • Every inspection is logged to an immutable audit trail
Powered by Veea Lobster Trap DPI engine
🤖
Step 4

AI Governance Analysis

Google Gemini powers the AI assistant that can analyze governance data, generate compliance reports, detect anomalies in spending patterns, and recommend policy improvements. It's like having an expert governance analyst available 24/7.

  • Analyzes enterprise governance posture and risk factors
  • Generates structured findings with severity ratings and evidence
  • Reviews financial data for anomalies and compliance gaps
  • Provides recommendations with specific remediation steps
Powered by Google Gemini 2.5 Flash
🏛️
Step 5

Democratic Governance

Citizens can claim governance roles (council member, minister, president), join institutions, vote on proposals, and participate in democratic decision-making. All governance actions are recorded on a transparent ledger that anyone can audit.

  • Claim a council seat in your city and participate in local governance
  • Join institutions and contribute to policy development
  • Vote on proposals using your verified civic identity
  • Every governance action is recorded with a SHA-256 audit hash
User-driven, identity-verified
🔍
Step 6

Transparency & Audit

Everything is auditable. Every score has a source. Every AI interaction is logged. Every governance decision is recorded. The platform is designed so that no one — not even administrators — can manipulate data without it being detected.

  • Score Ledger: See exactly which data points produced each score
  • Source Attribution: Every metric links back to its original data source
  • Audit Chain: SHA-256 hashed records that can't be altered retroactively
  • DPI Log: Complete history of AI security inspections and their outcomes
Built into every layer of the platform

What Populates What

Glossary of Terms

Plain-language explanations of technical terms you might encounter on the platform.

DPI

Deep Prompt Inspection — AI security that scans text for threats before and after it reaches the AI model

Lobster Trap

Veea's AI security framework that catches malicious inputs like a lobster trap catches unauthorized access attempts

Score Ledger

The complete, auditable record of how a score was calculated, including all data sources and computations

Pillar

One of the four major categories in the scoring formula (Economic, Social, Stability, Resources)

Trust Tier

How reliable a data source is: Tier 1 = official government, Tier 2 = international organizations, Tier 3 = commercial providers

PrismAPI

A financial data service that provides real-time economics data — interest rates, inflation, trade statistics

Compliance Pack

A set of security rules designed for a specific industry (HIPAA for healthcare, SOC2 for technology, Finance for banking)

Governance Score

A number from 0-100 measuring how well an entity (country, company, council) performs across multiple verified metrics

Civic ID

A verified digital identity created through official channels (national eID, passport verification, biometric check)

Audit Chain

An unbreakable sequence of records where each entry's fingerprint (hash) includes the previous entry, making tampering detectable