Wakama Score · core engine
An agricultural file your committee can read, point by point.
Wakama Score brings together the data on a farmer and their plots, checks that it is consistent and computes a score in which every contribution is sourced. Your analysts see what supports the file, what weighs on it and what still needs confirming.
The score supports analysis. It is neither a probability of default nor a credit decision.
Wakama Score
0/100
Grade
E
Contribution balance
Favourable factors push the score right of the base, watch points pull it left. Select a factor to see its detail.
Watch points
0
Base
50
Favourable factors
0
50 +0 −0 =
50/100
Partial score
No factor selected.
Synthetic file. The illustrative score of 72/100 is an analysis aid: it is neither a default probability nor a credit decision.
- Sector baseline50
- 3-season NDVI stability+9
- Cooperative history+8
- Buyer contract+7
- Verified geometry+5
- Rainfall deficit−4
- Yield to confirm−3
File, farmer, contributions and weights are fictitious, shown for illustration (final score 72). The actual methodology is presented in a meeting.
01 /Who it is for
Built for teams who assess agricultural credit.
Banks, MFIs and financial cooperatives lending to farmers, cooperatives or agricultural SMEs, who must justify every file before a committee.
- 01
Credit analysts
A structured file, readable contributions and a clear list of points to check.
Assessment
- 02
Credit committees
A short, sourced note that separates what is established from what still needs confirming.
Decision
- 03
Risk and compliance
An audit trail: which data, from which source, on which date, for which calculation.
Control
02 /Inputs and deliverables
Mixed data in, a readable file out.
Sources vary from one file to the next. Every data point keeps its origin, and the score states what it actually relies on.
Inputs, as available
- Farmer identity and details
- Plot geometry surveyed in the field
- Satellite imagery and vegetation indices
- Cooperative history and any buyer contracts
- Field agent observations and photos
- Weather station data where it exists
Deliverables
Explainable score
A score out of 100 and its grade, broken down into positive and negative contributions.
Consistency checks
Area, crop, yield, photos, duplicates: each check with its result.
Points to confirm
What must be verified before the decision, and why.
Committee file
An exportable summary with its sources, ready for internal review.
03 /Reading a score
A score is only worth something if it can be challenged.
Wakama Score is designed to be questioned by an analyst, not accepted as is.
- 01
Visible contributions
Each factor shows its weight and direction. Nothing is hidden in an average.
- 02
Attached sources
Each contribution points to the data used, its origin and its date.
- 03
Written uncertainty
Missing or weak data is flagged as such instead of being smoothed over.
04 /Roles and limits
Wakama informs the file. Your institution decides.
The score replaces neither your credit policy, nor your committee, nor the human review of the file.
What Wakama does
- Structure the data in an agricultural file
- Check consistency and flag gaps
- Compute an explainable, sourced score
- Prepare a summary for the committee
What your institution keeps
- Credit policy and eligibility criteria
- Lending decision and terms
- Client relationship and disbursement
- Regulatory responsibility
Limits and dependencies
- The score depends on the data available for each file
- Satellite imagery is sensitive to cloud cover
- Weather stations and drone missions are not available everywhere
- No default prediction is claimed
05 /Getting started
Start small, without heavy integration.
You can start without connecting to your information system. Any integration is scoped later with your teams.
- 1Meeting
Demo
See the method on fictitious files close to your cases, in a short meeting.
- 2Deliverables
Sample files
Analyse a few representative files and compare Wakama’s reading with yours.
- 3Tailored
Scoped perimeter
Define a value chain, area or portfolio, then the expected deliverables.
Next step
See Wakama Score applied to files close to yours.
A short demo, with no commitment, on fictitious data. We then discuss what would be useful for your teams.