Less money, more paperwork. One of those you can fix.
Grants, reports and outreach with a small team.
Priced in pula, rand and dollars · English and French · Built by OrionX in Botswana
Aid crosses oceans without difficulty. It is the last desk of paperwork it cannot cross.
The money went back to 2015. The obligations did not.
Official development assistance in 2025 — the largest single-year contraction on record
OECD DACBilateral aid to sub-Saharan Africa — a proportionate share of a very large cut
OECD DACBilateral grants: the flexible money that pays for staff, systems and rent
OECD DACThe year total aid has returned to — the year the SDGs were adopted
OECD DACIn 2025, official development assistance fell 23.1% to USD 174.3 billion — the largest single-year contraction on record — with bilateral aid to sub-Saharan Africa down 26.3%. The shape of the cut matters as much as its size: bilateral grants, the flexible money that pays for staff, systems and rent, fell 29.1% — faster than the total — so the average surviving grant is more restricted than a year ago.
More compliance work per dollar, less core funding to do it with, and a further 5.8% decline projected. Total aid is back at its 2015 level, the year the Sustainable Development Goals were adopted. The obligations are 2030’s. The money is 2015’s.
Read carefullySub-Saharan Africa’s 26.3% cut sits against a 26.4% bilateral average. Nobody singled the region out. It simply took a proportionate share of a very large cut — and a proportionate share is fatal to an organisation with no reserves.
Nor is the shock evenly spread across funders. US assistance fell 57% in a single year; Japan’s fell 6% (OECD DAC). An organisation with one donor is not carrying programme risk — it is carrying a portfolio position, and no amount of field performance hedges it.
US assistance, in a single year
OECD DACJapan’s assistance, over the same year
OECD DACNobody will fund the thing that makes you fundable.
Four steps that feed each other. The loop is the argument — and Uhuru does not add money to it. It removes one edge. Drag the model to see where the ring is cut.
The teal bar is the cut edge · drag to orbit · download the model as OBJ or GLB
The back office is the one part of the work with no budget line.
A funder will pay for the borehole and not for the hour spent writing down that the borehole exists. In a 20-organisation US study, true indirect costs ran at a 40% median while funders reimbursed 15% — and UNHCR allows local partners 4%.
You need good systems to win the funding that would have paid for good systems.
Same organisations. Same countries. Radically different throughput.
The money that survives still struggles to reach African organisations. Donors committed to placing at least 25% of humanitarian funding with local and national actors; in 2023 they managed 4.4% overall, and just 0.6% that can be independently verified as direct.
Yet the UN’s country-based pooled funds route 30% of their own allocations to those same organisations. Local capacity is plainly not the bottleneck — the documentation interface between a donor’s grant machinery and a local back office is.
Percentages of different bases — shown side by side, never divided. 4.4% and 0.6% are shares of total humanitarian funding; 30% is a share of the pooled funds’ own allocation envelope. Development Initiatives
The destination is identical. The difference is entirely in what happens on the way — and what happens on the way is paperwork.
The only exit that does not require a donor to change policy is a back office that costs almost nothing.
Uhuru is that back office, as working tools rather than a platform to administer. U Docs takes field notes, activity data and photographed flip charts and drafts the donor-templated report — narrative, logframe, indicator tables — in the format each funder demands, so the same quarter’s work can be written up for more than one donor.
That throughput matters because the shock is concentrated, not general: a single-funder NGO carries a risk no field performance can offset — and the cure is writing credibly to many donor formats at once. Helios, Uhuru’s frontier reasoning, works the hard cases: a logframe that has drifted from reality, a budget realignment memo, an evaluation response. WeDraft, built for African jurisdictions, prepares the MOUs, sub-grant and partnership agreements that otherwise wait weeks on a lawyer’s availability.
The evidence on who gains most from these tools favours your staffing shape. In the studies run so far — global, not yet African — AI assistance lifted newer workers’ output far more than experts’: 34% for novices in one large study against a 14% average, and 43% versus 17% in a field experiment with over 700 consultants. African NGOs are typically a wide junior base under a thin senior bench: exactly where those gains land largest.
The same research carries a warning we would rather tell you than have you discover in an audit: outside the model’s competence, assistance made skilled people worse. So Uhuru draws the line where the evidence draws it. Everything lives in Uhuru Cloud, in one place your auditor can see, and Uhuru Voice lets field officers file in speech what they would never have typed.
The 34% figure is customer-support work, not document drafting; the overall average in that study was 14%. NBER · MIT Sloan / Harvard Business School
We would rather tell you where the edge is than have you find it in an audit.
The model’s work
- Narrative
- Translation
- Drafting
- Formatting
Stays human, stays signed
- Beneficiary verification
- Financial reconciliation
- Attribution claims
Eight things that change in the first week.
Reports in the donor's format
U Docs drafts narrative reports, logframes and indicator tables from raw field material — in each funder's own template, ready for your edit rather than your weekend.
Proposal throughput
Helios and U Docs let a small team write credibly to many donor formats at once. Diversification is the honest answer to a 57% single-donor collapse — and diversification is a throughput problem.
Agreements without the wait
WeDraft prepares MOUs, sub-grant and partnership agreements for African jurisdictions. A sub-grant that used to hold up a quarter's disbursement now moves at the speed of the programme.
Two languages, one record
Uhuru Chat moves your work between the language you serve in and the language you report in. The translation gap African NGOs actually have is with their funders, not their communities.
A specialist for the big briefs
Bio-D, the biodiversity and climate specialist, for conservation and adaptation programmes — a model shaped for the briefs African NGOs actually run, not generic chat.
Filed by speech
Uhuru Voice lets a field officer dictate the visit on the walk back instead of typing it up at the end of a fourteen-hour day. The record the sector loses most is the one nobody had the energy to write.
One account your auditor can open
Every report, budget, agreement and version lives in Uhuru Cloud. An audit becomes a login rather than a reconstruction across three inboxes — and the institutional memory stops walking out with whoever leaves.
The line we will not cross
Narrative, translation, drafting and formatting are the model's work. Beneficiary verification, financial reconciliation and attribution claims stay human. The research is clear that outside its competence, assistance makes skilled people worse.
Eight lines a programme lead can check against her own week.
Bana ba motho ba kgaogana tlhogo ya tsie.
Children of the same mother share the head of a locust.
When there is almost nothing left, family shares even that — and wastes none of it. Aid is back at its 2015 level while the need is 2030-sized. The discipline the moment demands is not doing less; it is making the administration around the work cost almost nothing.
The programme lead who spends the week in the field and the weekend converting it into the formats that keep it funded.
Runs three districts on two grants, delivers well, and loses every second Saturday to the narrative report nobody else on the team can write.
Works in French with the community and in English with two of the three funders, and has priced a translator into a budget exactly once, successfully, in six years.
Fills three notebooks a season, types up perhaps half, and knows that the half he loses is the half that would have proved the programme worked.
Start with one report.
Bring the field notes from the quarter you have already delivered and see what comes back. No implementation, no migration, no consultant.