AgentBrain Research

AMR evidence architecture.

Research notes on the connection layer between policy, evidence, capital, execution and outcomes.

AgentBrain governance · 20 Aug 2026

How this intelligence was produced.

Work can be delegated without delegating canonical truth. This note explains how AgentBrain separates agent execution, evidence review, controlled reuse and human publication approval.

Read governance note
Capital traceability · 19 Aug 2026

What a traceable AMR funding chain looks like.

Switzerland's NRP 72 shows how an AMR mandate, CHF 20m programme, identifiable projects and synthesis outputs can form a reconstructable capital chain without overstating national outcome causality.

Read research note
Evidence discipline · 18 Aug 2026

When the evidence stops: why not found is an AMR finding.

Failure to locate evidence is not proof that an activity, budget or outcome does not exist. In AMR intelligence, unknown has to remain an evidence state.

Read research note
Health financing · 17 Aug 2026

Africa's health financing reset: what it means for AMR.

Ahead of WHO AFRO RC76, a new regional health-financing strategy puts domestic financing, accountability and country ownership at the centre. AMR is a useful test of whether those principles can become traceable implementation.

Read research note
Evidence connection · 17 Aug 2026

AMR does not have one data problem. It has a connection problem.

The global AMR response is generating more data than ever. The harder question is whether policy, financing, implementation and outcomes can be connected into a defensible evidence chain.

Read research note