Answer any question about your book on the day it is asked.
Arrears by origination quarter, exposure by state against the facility cap, which borrowers move as one obligor. Today each of those is a week of your finance team. On a connected book each is a query, rebuilt every fifteen minutes from your own systems, and the fund reads the same screen you do.
One state holds a fifth of the book, past the limit in the facility
Commerce and services carry two thirds of the book
The book grew 40% while arrears fell for two months running
9.5% of the book is unmeasured, not delinquent
Working with
Four readings a committee argues about.
Each is computed from the connected book rather than submitted, and each carries the loans that produced it. Nothing here asks the lender to change how it originates or services a loan.
One borrower in ten is unmeasured, not delinquent, and priced as if it were both.
Grey is the zone that costs a lender pricing. Those borrowers pay on schedule but reach no bank feed, so a fund has to price them as unknown until the feed reconnects. Separating grey from red is the single change that moves a spread.
Why those 173 are grey
118.9M, none of it lateGrey is not a softer red. A grey borrower is current and unmeasurable, and pricing the two the same is what a fund does when nobody separates them. Three of the four rows above are a connection problem, not a credit one.
Every reconnected feed moves exposure out of grey.
Grey fell from 13.5% to 9.5% in six months without a single change to credit policy.
One state can hold a fifth of the book, past the limit in the facility.
Concentration is a covenant, not a preference, so it is drawn against its limit rather than reported as a number. Guanajuato is 1.4 points outside.
The book grew 40% while arrears fell for two months running.
Growth with falling arrears is the shape that prices a facility. Growth with rising arrears is the shape that reprices it, and both look identical in a summary.
Each cohort since the scorecard change ages better than the last.
Read down a column, not across a row: at three months on book the newest cohort sits a third below the oldest.
Every one of them answers a question a committee asks out loud.
None of the four is a score. Each one carries the loans behind it, so the answer can be argued with rather than believed.
We do not replace the audit.
An audit reads a closed period once a year, and by the audit date the file is in order.
No auditor has the infrastructure to open every application when there are hundreds of thousands of them. The eleven months in between belong to nobody.
Signs the year once it is over.
That signature is the one thing we cannot give you, and we are not trying to. It closes a period that has already happened.
Once a year, on closed quartersHolds the year up to the light while it runs.
Which is the only window in which anyone can still act on what it shows, and the reason an investor keeps lending.
Every day, up to yesterdayOne reading, four people who need it.
The same connected book answers the lender, the warehouse bank, the fund behind the facility and the credit committee, which is what stops four versions of the truth from existing at once.
Find the obligor before the payment is missed
Three borrowers sharing a legal representative and a registered address are one exposure against a single obligor limit, found on a registry change.
Test the borrowing base between drawdowns
Eligible collateral and every covenant recomputed from the borrower tape rather than submitted at the drawdown, so a breach is found while a cure period is open.
Watch four lenders on one screen, at no cost
The lender pays for the platform. The fund gets the cabinet, the monthly report and an alert the day a book moves.
Every exception with an owner and a clock
An exception is only useful if someone owns it and it expires. Each carries the person accountable and the date the cure period ends.
A measured book is a cheaper book. Every time.
None of these is a promise about credit quality. They are consequences of the same change: the investor stops pricing what it cannot see.
The unknown share shrinks, and the spread follows.
Reconnecting a feed moves borrowers out of grey without a single change to credit policy. Figures illustrative.
Covenants stop arriving as surprises.
Two of four limits are outside and both are curable, which is a different conversation from finding them at the next quarterly certificate.
The second facility is easier than the first.
Twelve months of visible repayment behaviour is worth more at the next raise than any deck, and it exists only if somebody was watching.
The fund stops asking for an update.
A monthly reading that arrives without chasing removes the request both sides find most expensive.
Four books become comparable.
The same checks on every lender is what lets a fund rank two of them rather than argue about two formats.
Written for an investment committee, not for an engineer.
The same document the lender sees, so there is nothing to reconcile between two versions of the month. It lands on the first business day, and the delivery is on the record.
Portfolio
Outstanding balance, cohorts by vintage, roll rates and collections, rebuilt from source records rather than typed into a template.
Movement
What changed since last month and which loans caused it, so a trend can be read rather than guessed at.
Findings
Use of funds, related parties, concentration and transfers before a cut-off, each carrying the positions that triggered it.
Headroom
Borrowing base, eligibility and covenant distance recomputed as new data arrives, not once a quarter.
The report writes itself, which is the point.
Three of three, on the first business day.
The desks the lender approved, and nobody else.
Analytics on anonymised data, because the law requires it.
Data protection law in every market we operate in stops borrower level personal data from travelling to a third party for someone else’s analysis, and the product is built so that it never needs to. Identities stay inside the lender, and the lender remains the party that decides what may be shared at all.
Aggregation before delivery
Reporting is assembled from portfolio level figures, not from borrower files.
Processing on instruction
We act under contract on the lender's instruction rather than as an independent controller.
Scope agreed in writing
Fields, recipients and retention are fixed before a connection opens, and can be narrowed later.
Revocable at any time
The lender can close the connection, and reporting stops with it.
Aggregated at portfolio level before delivery. The lender can close the connection and reporting stops with it.
What we hold ourselves to.
A lender is handing us the most sensitive asset it owns. These are the terms, and each one is visible in the product rather than only in a policy page.
Read only, no write path
There is no route by which we can originate, amend, reprice or move a loan. The connection is one directional by construction.
Separation between lenders
No lender's data is used to build a report for another. Benchmarks, where shown, are aggregated across the book and never traceable to a counterparty.
Audit trail on access
Every read is logged with who, what and when, and the log is available to the lender rather than only to us.
Deletion on exit
When a lender leaves, the connection closes and retained data follows the schedule agreed in the contract.
We are not a bureau
We hold no credit bureau licence and do not score consumers. We read a lender's own portfolio with that lender's consent.
Every delivery, with the desk that received it.
| Recipient | Period | Sent |
|---|---|---|
| Andesa Credit Partners | 2026-07 | 01 Aug |
| Northbay Private Credit | 2026-07 | 01 Aug |
| Sierra Alta Capital | 2026-07 | 01 Aug |
Six of seven sources live, and none of them can be written to.
A source that cannot be written to cannot be used to change a loan record, which is the structural version of the promise rather than the contractual one.
Every figure, counterparty and borrower on this page is demonstration data created for illustration. It is not a real portfolio and not a real transaction.
Let us read one month of your book.
Connect a copy of the last twelve months and we will send back the report an investor would receive, including the findings, before you decide anything.
- ✓A real monthly report on your own data
- ✓The findings an investor would raise, before they raise them
- ✓No obligation to continue afterwards