Data engine
Ingestion, validation and reconciliation of the loan book, with exceptions surfaced before they reach the calculation.
Risk BI Platform
Your IFRS 9 close, in a system instead of a spreadsheet.
One governed platform for expected credit loss, stress testing, ICAAP and ILAAP — calibrated to how your regulator actually supervises you.
The problem
A spreadsheet is not a system.
The Risk BI Platform runs your entire IFRS 9 process as a governed system, with stress testing built into the same engine. The same inputs produce the same result, on demand, months later. Every figure traces back to source data through a visible calculation path. And the process survives the departure of whoever built it.
You already know what your ECL process looks like. A chain of linked workbooks. One person who fully understands it. No version control, no calculation lineage, and an audit trail that has to be reconstructed after the fact rather than produced on request.
So the reporting window is spent rebuilding the model rather than interpreting the result. Your auditor asks how a staging decision was reached and the answer takes three days to assemble. Your regulator queries a number and the honest response is that it will be addressed next quarter. And the whole apparatus depends on one person who could resign.
Then the same team, on the same foundation, is asked to produce ICAAP, ILAAP and a monthly or quarterly supervisory stress test.
None of this reflects the quality of your modelling. It reflects the fact that a spreadsheet is not a system.
Platform
Modules
Probability of default
PD estimation and calibration, term structures, and through-the-cycle to point-in-time conversion.
Loss given default
LGD estimation with recovery and collateral treatment.
Exposure at default
EAD including credit conversion factors for undrawn commitments.
Staging
SICR assessment and stage allocation, with the trigger logic visible and versioned rather than buried in a formula.
Effective interest rate
EIR computation and unwinding.
Macroeconomic overlay
Scenario definition, weighting and forward-looking adjustment to PD.
Stress testing
Scenario design and execution for ICAAP, ILAAP and periodic supervisory stress tests.
Reporting
Dashboards, disclosure tables and the reconciliations your finance team needs to close.
The part your auditor cares about.
Reproducibility.
The same inputs produce the same result on demand, months after the fact.
Calculation lineage.
Every number traces back to source data through a visible path. When an auditor asks how a figure was reached, the system answers.
Model versioning and change control.
Every methodology change is recorded, dated and attributable, with maker-checker controls on the changes that matter.
Continuity.
The process does not live in one person’s head or one person’s workbook. A team can run it, and a new joiner can learn it.
Performance
Speed, and what it unlocks
An ECL run that typically takes two to three days on a spreadsheet model completes in hours.
The time saved is the smaller half of that. What actually changes is that re-running becomes practical. You can test a staging assumption, re-run under a different macroeconomic scenario, answer a supervisor’s question or respond to an audit challenge within the same day — rather than committing to address it in the next reporting cycle. You stop running one scenario because one is all there is time for.
Implementation
Deployment and integration
Deployment.
On-premise, within your own network. This is how our clients run it and what we recommend. A hosted option is available where an institution prefers it.
Architecture.
Web application on a standard Microsoft stack — Internet Information Services and SQL Server. For smaller institutions both components can sit on a single virtual server.
Integration.
APIs are available for connection to most core banking platforms.
Implementation.
Typically six to eight weeks from contract to live use.
Support.
From Nairobi and Kigali, by the team that built the credit logic.
Your data stays in your bank.
The platform runs inside your network. Customer-level information remains entirely in your custody and within your jurisdiction, which is what supervisors in this region expect. We do not receive individual borrower records.
Evidence
Proof
Banks, microfinance institutions, SACCOs, consumer credit providers and insurers that report under IFRS 9 and are supervised under a Basel-based capital regime. It is scoped and priced for institutions that cannot justify a global vendor implementation, and it scales down as readily as up.
18 deployments across 13 institutions and groups, in Kenya, Uganda, Tanzania, Botswana, Zambia and Barbados.
Running at commercial banks, microfinance banks and institutions, consumer credit providers and a reinsurer — the same system, across every tier of institution.
Four client groups run it in more than one country. One took it into a new market on another continent as part of that market’s launch.
Expected credit loss output produced on our platform has been examined by all four Big Four audit firms and by BDO, in the course of their audits of institutions running it.
Next step
See it running on your own portfolio.
A demonstration takes an hour and uses your data. We will show you the calculation lineage, the staging logic and the audit trail — the three things your auditor asks about first.