These are the principles the platform is held to, at the level a diligence reviewer would want to hear them, rather than as a methodology walkthrough.
Bring together conflict, climate, governance and land-and-water data, and build it locally where no reliable feed exists.
Resolve everything to the district and the site, not the country.
Read each district against the rest of its country, so a score means “relative to here”.
Attach confidence, the evidence behind it, and the date it holds for. Thin evidence is flagged, not hidden.
Turn the reads into conditions and monitoring triggers a decision can use.
Open any score and see what is inside it: the inputs, how the district compares, and how much each one moves the number.
How this score is computed
Governance · Example asset 1, District A: 41
Accountability
Protest and grievance activity
Infrastructure
Roads, services and investment
State presence
Reach of public services
Confidence and ranges travel with the number, so thin evidence is never smoothed into a single confident reading.
Drought outlook · Example asset 1, District A
The point comes with its range: the outlook could sit anywhere inside the whiskers.
Change is reported only between real data refreshes. A forecast carries its target date, and an outdated one says so.
Conflict · Change since last refresh
52 → 58, rising
Two refreshes compared. A change, not yet a trend.
Climate outlook · Expired
This outlook targeted an earlier season. It is shown for reference until the next refresh, not as a current forecast.
Every read looks both ways: the local dynamics that could affect the investment, and what the investment could mean for the people around it.
On request, alongside a demo:
Our methodology draws on peer-reviewed work in sustainability risk assessment and analytics.