Methodology

How we calculate scores, generate briefs, and verify partners.

The Three Scores

Every country is assigned three composite scores updated daily from verified data sources. Each score runs from 0 to 100.

Need (0–100)

Composite of humanitarian pressure, food insecurity, health burden (under-5 mortality, disease prevalence), and displacement. Higher = greater need.

Opportunity (0–100)

GDP growth trajectory, internet and mobile connectivity, startup ecosystem activity, trade openness, and youth demographics. Higher = more opportunity.

Stability (0–100)

Governance quality (V-Dem), conflict pressure (ACLED events), rule of law, institutional performance, and political stability indices. Higher = more stable.

Data Sources

All indicators are sourced from official, peer-reviewed, or institutionally verified datasets. No proprietary or unverifiable sources are used.

World Bank Open DataGDP, poverty, education, infrastructure
UN SDG Indicators17 goals, 230+ indicators
WHO Global Health ObservatoryMortality, disease burden, health systems
ACLEDArmed conflict events, fatalities, protest data
GSMA IntelligenceMobile connectivity, internet access
V-Dem DatasetGovernance, democracy, institutional quality
OCHA ReliefWebHumanitarian situation reports
AfDB African Economic OutlookMacroeconomic forecasts, regional analysis

AI Briefs

AI briefs are generated using a Retrieval-Augmented Generation (RAG) pipeline. The model is given only the current day's ingested data — it cannot draw on prior memory or training data to make factual claims.

  • •Every factual claim in a brief must link to a source in the citation list.
  • •Briefs without citations are never published.
  • •The confidence score (0–1) reflects citation coverage and source quality.
  • •Stale briefs (>48h) are flagged and the last valid snapshot is shown.

Partner Verification

Organisations listed in the Partner Directory are reviewed by our editorial team on a rolling basis.

A

Independently audited financials, published impact reports, and at least one third-party evaluation.

B

Self-reported financials reviewed by our team, clear mission alignment, active programme evidence.

C

Listed but not fully reviewed. Proceed with independent due diligence.

Freshness & Degradation

Each data point carries a freshness flag: fresh (updated within 24h), aging (24–72h), or stale (over 72h or source unavailable). On source failure, the platform returns the last valid snapshot and displays a freshness warning rather than showing empty data.