AR
Credit risk assessment for receivables
Where is your future money concentrated, and what are the risks? Who settles on time, whose lateness is widening, and how much of your financial accounts hangs on a single name.
Decide on evidence, not instinct
The credit risk data engine
Financial data, properly read — so a default never arrives unannounced
A lot of invoices across a lot of customers, and the picture is blurry: you cannot see where the exposure sits. The engine reads them invoice by invoice, scores every counterparty with the reasons behind the number, and turns the pile into a clear view, so your decision stands on evidence.
The idea
Your financial accounts know — nobody has read them
Invoices arrive as one slab: every one the same width, the same height, nothing to tell one counterparty from another. Reading is what pulls them apart. The high-risk head separates and shrinks to the share of the money it actually is, and the bulk settles into the low band.
Risk ↑ Higher
Lower
Illustrative financial accounts of 12 invoices, SAR 8,620,000 of exposure in total. Once read, 8% of the value sits in the high band, 20% in the medium band and 72% in the low band. 3 of the invoices did not pass the document check.
Invoice with no valid stamp, or a suspected duplicate
Column width is its share of the financial accounts by value
| Band | Invoices | Exposure | Share |
|---|---|---|---|
| High risk | 3 | SAR 730,000 | 8% |
| Medium risk | 3 | SAR 1,710,000 | 20% |
| Low risk | 6 | SAR 6,180,000 | 72% |
| Total | 12 | SAR 8,620,000 | 100% |
The figures in the drawing are ours, written to show how the reading works. They are not Fkah's financial accounts and not any client's.
The engine at a glance
The buckets are labelled; no counts are shown
Scattered invoices
Invoices from different customers arrive apart, and give no clear picture.
Mudarhim gathers them
Mudarhim is the engine: it reads the invoices and the statement, matches the stamp, knows the one recorded twice, and works out how long each has run.
The buckets
The invoices settle into labelled buckets: the financially sound, the ones that need a look, and the ones documented and matched.
A clear view
Out comes a read for each customer: their invoices and how long they have run, the reasons behind the read, and what changed. The analysis supports the decision, and the decision stays with the reader.
An illustration. The customer names here are our own example, written to explain how the read works; they are not Fkah's accounts or any customer's accounts.
The method
How the engine reads your financial accounts
Four steps, in the same order every time, across receivables and payables alike.
It reads the whole of the financial accounts
Every invoice in both directions: what your customers owe you, and what you owe your suppliers. No sample, and not just the largest ten counterparties.
It verifies the document
It checks the ZATCA stamp and the hash chain, and flags anything that looks like a duplicate inside the same financial accounts. Your money, beyond dispute: the stamp proves the document is authentic and the seller's sequence intact. It proves nothing beyond that.
It scores the counterparty
Payment behaviour, length of dealing, and the share of the financial accounts a single counterparty carries. Each finding is a declared reason, and each reason carries a fixed weight.
It explains the score
The explanation and the score are one piece of arithmetic: the reasons you read are the reasons that made the number, so the two cannot disagree.
The engine reads, scores and warns, and supplies analysis that supports your decision. The decision stays yours.
The score
A score you can read line by line
It opens at 60: a counterparty nobody has read anything about is not safe, it is unread, and it has to earn its way up. Each reason then moves the number up or down, and the final score is the sum of them.
An illustration, using the engine's own weights. A reason carries the same weight wherever it appears, so two counterparties with the same finding are read the same way.
Each line is one finding about the counterparty, with what it did to the number beside it.
Opening reading60
- Every invoice carries a valid ZATCA stamp+8
- More than three years of continuous dealing+7
- Days beyond terms widening quarter on quarter-22
Score53
The distinction
A government buyer is read differently
A government body settles well beyond the agreed terms. The question there is when the cash lands, not whether it lands.
The reading that confuses the two
Everything past its due date is counted into the at-risk figure, so the largest government payer in your financial accounts tops the danger list while being the soundest thing in it. A tool that teaches you that is teaching you the wrong thing.
How the engine reads it
Government payment follows the budget cycle, not capacity to pay, so it reads as cash timing and is kept out of the at-risk figure deliberately. The timing shows up where it belongs: in days to collect.
The solutions
Two directions, one reading
The same reasons and the same score, applied once to what you are owed and once to what you owe.
AP
Credit risk assessment for payables
The risk on your suppliers: who you depend on, how much of it you owe, and what changes in your operation if one of them fails.
Who it is for
Two audiences, the same reading
One engine, two questions. A company that wants to know its own risks, and a financial institution that wants to know its client's risk before it decides.
For companies
Know your risks: a score on every counterparty in your financial accounts, with the reasons that built it. Analysis that supports your commercial judgement rather than dictating it, so you see where the exposure gathers before it becomes a surprise.
For financial institutions
Know your client's risk before financing: the engine reads your own portfolio and supplies the analysis and evidence a credit decision rests on — a verified document, readable payment behaviour, an explained score. The decision stays with you alone.
Next step
Tell us about your financial accounts
Send us your details and we'll get in touch within two business days. No commitment, and you can opt out whenever you want.
Your data
- We work to SDAIA rules, and the trust page says exactly where the data sits today
- A published privacy notice covering what we collect and where it is held today
- Your financial accounts are analysed for you alone. We do not pool client financial accounts, and we never build a score about one company out of another company's data.
Questions
Frequently asked questions
What is the credit risk data engine?
An engine that reads your receivables and payables accounts: it verifies every invoice, scores each counterparty, and shows where the score came from. The engine is software bought by subscription: it supports your decision and never takes it.
What is the difference between the receivables reading and the payables reading?
Receivables reads whoever owes you: your customers and how they pay. Payables reads whoever you owe: your suppliers and what you have committed to with them. Both directions use the same reasons and the same score.
How is this different from a credit bureau report?
The engine reads your own financial accounts: your invoices, your counterparties, and payment behaviour whose evidence you can see for yourself, line by line. A credit bureau report is a different thing, drawn from records that financial institutions share between them. The two answer different questions.