Which institutions and capabilities are in place, and how well are they documented?
01 · The problem
Do the documents answer your question?
Before a funding decision, you may already have the national plan, a budget and reports from ongoing projects. You still need to compare them to judge whether the proposed programme fills a gap.
We check which records refer to the same work and what they establish. Approved funding is kept separate from work delivered, and targets from measured results. Where public records are missing, activity on the ground remains an open question.
Your team can use the sources to review and explain its decision.
02 · Typical questions
What do you need to find out?
Where do funded programmes overlap? Which needs are still unmet?
After a funding commitment, what can be traced through procurement and delivery to the result?
Should the proposed programme be funded? Which assumptions need checking first?
Which planned activities were carried out? What is known about their effects?
Which investments and infrastructure projects affect this location?
04 · Applications
AMR programmes and location decisions.
Analytics can be used for questions about countries, portfolios, programmes or individual places. These examples show the kinds of sources involved in a public health question and a location review.
What follows national commitments?
For antimicrobial resistance (AMR), plans, surveillance data and funding records are held by different institutions. We examine what they establish about a programme and its delivery.
Method example. No customer results are shown here.
Explore the AMR use case
What reaches a place?
For an agreed area, we compare public investment and infrastructure projects with plans and local economic data. You can see which projects are announced or delivered, and where information is missing.
JaffnaKilinochchiMannarMullaitivuVavuniya
Discuss a place decision
05 · 5-Day Evidence Sprint
Start with a scoped source review.
The 5-Day Evidence Sprint reviews sources for a specific decision. With you, we define the question, area and period to examine, and agree which records to include. You receive an initial decision brief. It does not replace a full risk assessment, and funding decisions remain with your team.
Discuss a sprintA record of the agreed question, area, period and person responsible for the decision.
A list of the records reviewed and the findings each one supports.
Each main finding is marked as supported, partly supported or not established.
Connections between funding, projects and results, where the sources establish them.
Missing information and contradictory sources are identified.
A record of which additional sources or confirmations could change the answer.
06 · Who it is for
For teams that need to explain their decision.
You are reviewing a country or portfolio and need to assess whether the conditions for delivery are in place.
Before funding or expanding a programme, you need to know what is already funded and which needs remain unmet.
You need sources and findings that you can check and use in your own country or portfolio analysis.
You compare policy priorities with budgets and responsibilities, then examine delivery reports.
You need to assess infrastructure projects and unresolved risks before deciding on a location.
07 · Make limits visible
How well is a finding supported?
The sources determine how much weight a finding deserves, regardless of how confident an AI answer sounds. Each main finding is labelled to show what the reviewed records establish.
The reviewed sources directly support the finding or connection.
There is relevant evidence, but an important part of the connection is still unclear.
The available records do not provide enough information to answer.
Credible sources disagree on this point.
The reviewed records do not provide sufficient evidence for the claimed connection.
08 · Human review
Your team is responsible for the decision.
AgentBrain assists with research and comparing records. Your team reviews the findings using its own expertise and decides what conclusions to draw.
The brief distinguishes a source that suggests a connection from one that establishes it. In AMR, for example, health funding must be shown to belong to the programme. Planned results remain labelled as planned unless there is evidence they occurred.
SourceScopeStatusReviewDecision
09 · Stay traceable
Follow a finding back to its source.
When someone on your team questions a finding, they need to be able to check what it rests on. The review therefore records how the findings were reached.
The decision being prepared and the agreed scope are recorded.
You can see which record supports a finding and what it says.
Established connections are distinguished from those with only partial support.
Gaps and contradictions are documented for further review.
Recorded assumptions show what needs revisiting when new sources appear or programmes change.
10 · Your question
What do you need to decide?
Describe the decision you need to make and the records you already have. For example: should a programme receive funding in a region where similar work is already under way? That gives us a starting point for discussing the question and scope of the source review.
Your decision question
Records you already have
Your decision deadline