Adaptbase
Organizing the world’s climate adaptation knowledge.

Climate adaptation is atrillion-dollar challenge.

Today the world spends $190 billion a year defending 1.2 billion people to developed-economy standards. Protecting everyone exposed to climate hazards would cost $540 billion.

McKinsey Global Institute, 2025 →

We’ve cataloged data on 2,877 adaptation actions, 757 solutions, and 12,868 outcomes in 1,309 cities globally.

How will you use it to make better decisions about adaptation investments and innovations?

For Infrastructure investors
Infrastructure investor asks
FROM THE ADAPTBASE KNOWLEDGE GRAPH
Stormwater, wastewater and coastal protection lead the pipeline, at wildly different scales — the 18 projects reporting capex run from $2m to $15bn.
By asset class
  • Stormwater 19
  • Wastewater 14
  • Coastal protection 9
  • Water supply 8
93 capital projects, 62 with an asset class · as of September 2026
For Businesses
Business asks
FROM THE ADAPTBASE KNOWLEDGE GRAPH
Davao City leads with 10 flood-related actions targeting transport, ahead of Makati (6) and Edmonton (5).
Top jurisdictions
  • Davao City 10
  • Makati 6
  • Edmonton 5
  • Halifax 5
A facility-siting risk lens · 221 actions across 65 jurisdictions · as of September 2026
For Planners
Planner asks
FROM THE ADAPTBASE KNOWLEDGE GRAPH
Socio-Economic & Public Health Systems and Urban Ecology & Natural Capital draw the most dual-hazard investment — 228 and 221 actions respectively address both heat and flooding through the same system.
By urban system
  • Socio-Economic & Public Health Systems 228
  • Urban Ecology & Natural Capital 221
  • Hydrological & Water Infrastructure 169
  • Built Environment & Land Use 147
Where one investment pays off twice · 482 dual-hazard actions across 51 urban systems · as of September 2026
For Advocates
Advocate asks
FROM THE ADAPTBASE KNOWLEDGE GRAPH
The most-reported outcomes aren't risk metrics at all — they're public health (2,145) and social cohesion (1,268), ahead of ecological (709).
By benefit claimed
  • Public health 2,145
  • Social cohesion 1,268
  • Ecological 709
  • Service continuity 426
5,893 co-benefit claims from 1,012 actions · as of September 2026
How it’s built

Knowledge about effective adaptation is scattered across the web.

We built AI agents to collect millions of documents recording humanity’s adaptation plans and actions — and read them all.

CITY PLAN PDF · 2022 · PAGE 30
Miami-Dade County Extreme Heat Action Plan
ACTION 18: Pilot and scale cool pavements

Conduct pilots of different types of cool pavements in low-impact areas, such as walkways, bike paths, parking lots, and low-volume roads…

Cool pavement could either be permeable, allowing water and water vapor into the voids of the pavement that keep it cooler when moist, or it could be more reflective than standard pavement, reducing incoming solar radiation and cooling the air temperature.

We collect the facts that matter.

Using an ontology developed with adaptation and resilience experts, we extract structured evidence — so every document reporting on every city and company is read in a rigorous, standardized way. No chatbot guessing from a pile of PDFs.

PlanMiami-Dade County Extreme Heat Action Plan
ActionConduct pilots of different types of cool pavements
SolutionCool pavements
MechanismReducing incoming solar radiation
One shared vocabulary: 21 entity types, 60+ relationship types
Connecting the dots

Then we connect them.

One line in one city’s PDF becomes a fact you can count, compare and trace — against every other plan we’ve read.

prescribes works by implements SOLUTION Cool pavements PLAN Miami-Dade County Extreme Heat Action Plan MECHANISM Reducing incoming solar radiation ACTION Conduct pilots of different types of cool pavements
THE SENTENCE BEHIND THIS LINK · MIAMI-DADE, P. 30
“Cool pavement… could be more reflective than standard pavement, reducing incoming solar radiation and cooling the air temperature.”
The graph

Together, they form a knowledge graph.

0
entities
0
relationships
Deep research

This is only the start.
We deploy autonomous AI agents to expand, update and fill gaps in our knowledge.

EXPAND THE GRAPH
“Which startups are building technology-enabled solutions for extreme heat?” No city plan answers that, and we hold zero suppliers today — so the researcher leaves the library and works the open web.
→ new sources acquired
EXTRACT STRUCTURED DATA
“Build the financing timeline across every capital project we know of.” 93 projects, 163 financial instruments — the same graph, at far higher resolution.
→ queued
CONTINUOUSLY UPDATE AND REFINE
“Does anything that just landed change what we already hold?” Every document the researcher pulls is read against the whole graph, not only the task that fetched it. A Rotterdam report retrieved for a cooling-centre question revised a rooftop water-storage action we'd held as planned since 2021 — now 34 buildings delivered.
→ existing fact superseded
The foundation

Built to be built on.

WHAT COMES BACK
Facts, not paragraphs — each one still carrying the plan, the page and the sentence it came from. Your agent can show its work, and you can check it.
Any assistant · any workflow · over MCP
Research assistants
that cite real plans
City dashboards
that compare against peers
Funding tools
that show their sources
API  ·  MCP
Where we are
0
cities
0
documents
0
entities
0
relationships

Every relationship traces back to a quote: 104,419 evidence records.

Adaptbase
The knowledge graph for urban climate adaptation.

Explore our prototypes and publications

SCROLL