Creditcrest Enrich

Raw bank descriptions in; categories and payees out, each with its reason.

Part of Creditcrest Statement, in Creditcrest Decisioning Suite.

Every match has a reason.

A category and a payee for every transaction, with the rule that matched.

What you do here. Paste a list of transactions and get a category and payee for each, with the rule that matched.

Enrich transactions Try it end to end Take the tour

You bring: A list of transactions
You get: Category, third party, confidence, and the rule that matched
Holds the line: An unknown payee is flagged, never guessed

The fourth matches no rule and is returned uncategorised, not guessed.

An invented example

Nineteen invented transactions chosen to reach every confidence band, including the three that say the product does not know. Nothing below is anybody's data.

19 transactions, read

From the mixed example, which is invented. Engine 1.0.0, 201 rules from the AU merchant list. Read under the Balanced setting: The rules, the bank’s merchant code where a connection sent one, a register of Australian places, and a modelled reading where it keeps its category at 95% precision or better.

Category, by money: 79.0%
Category, by row: 84.0%
Third party, by money: 29.0%
Third party, by row: 47.0%

Coverage by money and coverage by row are two different questions and neither stands in for the other. One unrecognised two-thousand-dollar payment is a whole commitment missing from an assessment; forty unrecognised coffees are not.

How sure it is, and why

Band	Rows	Share	Money	Confidence	What it means
named	6 rows	32.0%	$645	99.0%	A rule matched the merchant’s name in the description.
kind	4 rows	21.0%	$4,503	99.0%	A rule matched a word for a kind of business inside a name this has never seen. It puts the money on the right line and it does not identify the payee.
paired	2 rows	11.0%	$1,160	95.0%	The holder’s own money moving between their accounts: an app’s split of a payment into a saver, or that share arriving in the saver the same day for the same amount, whatever name it carries. Not yet measured on labelled statements.
coded	1 row	5.0%	$15	90.0%	No rule matched the words. The bank sent a code that settles the line: the card scheme’s code for the kind of business, or its own word for a fee or interest. Not yet measured on labelled statements.
listed	1 row	5.0%	$18	99.0%	No rule matched the words. A public register names the business and says what kind it is: Wikidata by the web address on the line, a register of organisations (charities, councils, health funds, universities, chains) by their name, or a register of Australian places (Overture Maps, Foursquare and state registers) by these words in the suburb the line names. Measured on public registers and sample statements, not yet on lenders’ statements.
modelled	1 row	5.0%	$33	99.0%	No rule matched. A model fitted in this repository read the words; the confidence is the share of its answers at this score, in this category, that were right on lines held apart to measure it.
inferred	1 row	5.0%	$1,800	35.0%	No rule matched. The size and direction of the money were used instead, which is a guess and is marked as one.
declined	1 row	5.0%	$400	0.0%	The name is recognised and is deliberately not categorised, because the name does not settle what the payment was.
unknown	2 rows	11.0%	$373	0.0%	Nothing in the description matched a rule.

The confidence on each band is measured rather than asserted: what it gets right shows the run it comes from, the corpora it was run over, and the categories it is weakest at.

Every transaction, and why

Open any row for the rule and the words that matched it. A prediction whose reason cannot be read is not evidence.

Date	Description	Amount	Category	Third party	Confidence
1 Aug 2026	DIRECT CREDIT 019283 ACME PTY LTD PAYROLL	$2,320.00	Salary or wages	unknown	kind
Why this, and how sure A rule matched a word for a kind of business inside a name this has never seen. It puts the money on the right line and it does not identify the payee. Confidence 99.0% in the category. 185 of the 185 predictions this band made landed on an acceptable category when the rules were run over payee lists nobody here chose, on 5 November 2026. As given DIRECT CREDIT 019283 ACME PTY LTD PAYROLL Normalised DIRECT CREDIT ACME PTY LTD PAYROLL Words matched `PAYROLL` Direction money in Rule rule 86 in models/merchants/au.txt — it assigns `income_salary` and applies only to money in The words that matched name a KIND of business rather than a business, so the money is on a line and the payee is still unknown.
1 Aug 2026	Holiday Fund (25%)	−$580.00	Transfer	unknown	paired
Why this, and how sure The holder’s own money moving between their accounts: an app’s split of a payment into a saver, or that share arriving in the saver the same day for the same amount, whatever name it carries. Not yet measured on labelled statements. Confidence Nought, because no claim about a category was made. That is not a weak answer; it is the absence of one. As given Holiday Fund (25%) Normalised HOLIDAY FUND 25 Words matched `Holiday Fund (25%)` Direction money out Rule no rule matched
1 Aug 2026	ACME PTY LTD	$580.00	Transfer	unknown	paired
Why this, and how sure The holder’s own money moving between their accounts: an app’s split of a payment into a saver, or that share arriving in the saver the same day for the same amount, whatever name it carries. Not yet measured on labelled statements. Confidence Nought, because no claim about a category was made. That is not a weak answer; it is the absence of one. As given ACME PTY LTD Normalised ACME PTY LTD Words matched `Holiday Fund (25%)` Direction money in Rule no rule matched
2 Aug 2026	WOOLWORTHS 1248 RICHMOND AUS Card xx7113	−$142.86	Groceries	Woolworths `WOOLWORTHS`	named
Why this, and how sure A rule matched the merchant’s name in the description. Confidence 99.0% in the category. 499 of the 499 predictions this band made landed on an acceptable category when the rules were run over payee lists nobody here chose, on 5 November 2026. As given WOOLWORTHS 1248 RICHMOND AUS Card xx7113 Normalised WOOLWORTHS RICHMOND AUS CARD Words matched `WOOLWORTHS` Direction money out Rule rule 101 in models/merchants/au.txt — it assigns `groceries`
3 Aug 2026	SQ *WHITE HORSE COFFEE Sutherland NS AUS	−$8.40	Eating out and takeaway	unknown	kind
Why this, and how sure A rule matched a word for a kind of business inside a name this has never seen. It puts the money on the right line and it does not identify the payee. Confidence 99.0% in the category. 185 of the 185 predictions this band made landed on an acceptable category when the rules were run over payee lists nobody here chose, on 5 November 2026. As given SQ *WHITE HORSE COFFEE Sutherland NS AUS Normalised WHITE HORSE COFFEE SUTHERLAND NS AUS Words matched `COFFEE` Direction money out Rule rule 116 in models/merchants/au.txt — it assigns `dining` The words that matched name a KIND of business rather than a business, so the money is on a line and the payee is still unknown.
4 Aug 2026	NITHIKS KITCHEN ROZELLE 4421 AUS	−$63.00	unknown	unknown	unknown
Why this, and how sure Nothing in the description matched a rule. Confidence Nought, because no claim about a category was made. That is not a weak answer; it is the absence of one. As given NITHIKS KITCHEN ROZELLE 4421 AUS Normalised NITHIKS KITCHEN ROZELLE AUS Words matched nothing matched Direction money out Rule no rule matched
5 Aug 2026	Direct Debit 142619 TPG Internet DH9Q0HB22EP	−$69.99	Phone and internet	`TPG`	named
Why this, and how sure A rule matched the merchant’s name in the description. Confidence 99.0% in the category. 499 of the 499 predictions this band made landed on an acceptable category when the rules were run over payee lists nobody here chose, on 5 November 2026. As given Direct Debit 142619 TPG Internet DH9Q0HB22EP Normalised DIRECT DEBIT TPG INTERNET DH9Q0HB22EP Words matched `TPG` Direction money out Rule rule 94 in models/merchants/au.txt — it assigns `telco`
6 Aug 2026	RETAIL PURCHASE BP CONNECT 2214, MASCOT 1610	−$88.10	Transport and fuel	BP `BP`	named
Why this, and how sure A rule matched the merchant’s name in the description. Confidence 99.0% in the category. 499 of the 499 predictions this band made landed on an acceptable category when the rules were run over payee lists nobody here chose, on 5 November 2026. As given RETAIL PURCHASE BP CONNECT 2214, MASCOT 1610 Normalised RETAIL PURCHASE BP CONNECT MASCOT Words matched `BP` Direction money out Rule rule 105 in models/merchants/au.txt — it assigns `transport`
8 Aug 2026	WOOLWORTHS TEAM BANK	−$400.00	unknown	`WOOLWORTHS TEAM BANK`	declined
Why this, and how sure The name is recognised and is deliberately not categorised, because the name does not settle what the payment was. Confidence Nought, because no claim about a category was made. That is not a weak answer; it is the absence of one. As given WOOLWORTHS TEAM BANK Normalised WOOLWORTHS TEAM BANK Words matched `WOOLWORTHS TEAM BANK` Direction money out Rule rule 0 in models/merchants/au.txt — it assigns `uncategorised` This name is in the rules with no category against it, on purpose. It is recognised, and it is deliberately not categorised, so that no broader rule below it can guess — and so that nobody is sent to identify a payee somebody has already identified.
10 Aug 2026	OSKO PAYMENT REF 88213 J HARPER	$1,800.00	Other income	unknown	inferred
Why this, and how sure No rule matched. The size and direction of the money were used instead, which is a guess and is marked as one. Confidence 35.0% in the category — the categoriser’s own figure for a guess made from the size and direction of a credit rather than from a name. It is the one band no payee list can measure. As given OSKO PAYMENT REF 88213 J HARPER Normalised OSKO PAYMENT J HARPER Words matched nothing matched Direction money in Rule no rule matched
12 Aug 2026	AFTERPAY INSTALMENT 3 OF 4	−$37.25	Buy now pay later	Afterpay `AFTERPAY`	named
Why this, and how sure A rule matched the merchant’s name in the description. Confidence 99.0% in the category. 499 of the 499 predictions this band made landed on an acceptable category when the rules were run over payee lists nobody here chose, on 5 November 2026. As given AFTERPAY INSTALMENT 3 OF 4 Normalised AFTERPAY INSTALMENT 3 OF 4 Words matched `AFTERPAY` Direction money out Rule rule 61 in models/merchants/au.txt — it assigns `bnpl`
14 Aug 2026	SPORTSBET PTY LTD DEPOSIT	−$120.00	Gambling	Sportsbet `SPORTSBET`	named
Why this, and how sure A rule matched the merchant’s name in the description. Confidence 99.0% in the category. 499 of the 499 predictions this band made landed on an acceptable category when the rules were run over payee lists nobody here chose, on 5 November 2026. As given SPORTSBET PTY LTD DEPOSIT Normalised SPORTSBET PTY LTD DEPOSIT Words matched `SPORTSBET` Direction money out Rule rule 11 in models/merchants/au.txt — it assigns `gambling`
15 Aug 2026	RENT PAYMENT RAY WHITE BALMAIN	−$2,150.00	Rent	unknown	kind
Why this, and how sure A rule matched a word for a kind of business inside a name this has never seen. It puts the money on the right line and it does not identify the payee. Confidence 99.0% in the category. 185 of the 185 predictions this band made landed on an acceptable category when the rules were run over payee lists nobody here chose, on 5 November 2026. As given RENT PAYMENT RAY WHITE BALMAIN Normalised RENT PAYMENT RAY WHITE BALMAIN Words matched `RENT` Direction money out Rule rule 92 in models/merchants/au.txt — it assigns `housing_rent` and applies only to money out The words that matched name a KIND of business rather than a business, so the money is on a line and the payee is still unknown.
18 Aug 2026	AGL ENERGY LTD ELECTRICITY ACCOUNT	−$186.40	Utilities	AGL `AGL`	named
Why this, and how sure A rule matched the merchant’s name in the description. Confidence 99.0% in the category. 499 of the 499 predictions this band made landed on an acceptable category when the rules were run over payee lists nobody here chose, on 5 November 2026. As given AGL ENERGY LTD ELECTRICITY ACCOUNT Normalised AGL ENERGY LTD ELECTRICITY ACCOUNT Words matched `AGL` Direction money out Rule rule 47 in models/merchants/au.txt — it assigns `utilities`
20 Aug 2026	PHARMACY DEPOT HURSQPS HURSTVILLE AUS	−$24.95	Health	unknown	kind
Why this, and how sure A rule matched a word for a kind of business inside a name this has never seen. It puts the money on the right line and it does not identify the payee. Confidence 99.0% in the category. 185 of the 185 predictions this band made landed on an acceptable category when the rules were run over payee lists nobody here chose, on 5 November 2026. As given PHARMACY DEPOT HURSQPS HURSTVILLE AUS Normalised PHARMACY DEPOT HURSQPS HURSTVILLE AUS Words matched `PHARMACY` Direction money out Rule rule 113 in models/merchants/au.txt — it assigns `health` The words that matched name a KIND of business rather than a business, so the money is on a line and the payee is still unknown.
22 Aug 2026	MANNIX & CO 8821	−$310.00	unknown	unknown	unknown
Why this, and how sure Nothing in the description matched a rule. Confidence Nought, because no claim about a category was made. That is not a weak answer; it is the absence of one. As given MANNIX & CO 8821 Normalised MANNIX & CO Words matched nothing matched Direction money out Rule no rule matched
24 Aug 2026	RIVERSIDE BOOKSHOP 4421 ROZELLE AUS	−$32.50	Shopping	unknown	modelled
Why this, and how sure No rule matched. A model fitted in this repository read the words; the confidence is the share of its answers at this score, in this category, that were right on lines held apart to measure it. Confidence 97.4% in the category: the share of the model’s shopping answers at this score that were right on lines held apart to measure it. Given under the balanced setting; its score had to reach 0.5938. As given RIVERSIDE BOOKSHOP 4421 ROZELLE AUS Normalised RIVERSIDE BOOKSHOP ROZELLE AUS Words matched nothing matched Direction money out Rule no rule matched Model score 0.7575; the words that moved it most: `BOOKS`, `BOOKSHOP`
26 Aug 2026	SQ *MINTJAY 0042 BONDI AUS	−$14.50	Eating out and takeaway	`Mintjay`	coded
Why this, and how sure No rule matched the words. The bank sent a code that settles the line: the card scheme’s code for the kind of business, or its own word for a fee or interest. Not yet measured on labelled statements. Confidence 90.0% in the category, a stated judgement: the bank’s own code settled the line, and no labelled statement carrying codes has measured it yet. Given under the balanced setting. As given SQ *MINTJAY 0042 BONDI AUS Normalised MINTJAY BONDI AUS Words matched nothing matched Direction money out Rule no rule matched
28 Aug 2026	LUNE CROISSANTERIE FITZROY VIC AUS Card xx7113	−$18.40	Eating out and takeaway	`Lune Croissanterie`	listed
Why this, and how sure No rule matched the words. A public register names the business and says what kind it is: Wikidata by the web address on the line, a register of organisations (charities, councils, health funds, universities, chains) by their name, or a register of Australian places (Overture Maps, Foursquare and state registers) by these words in the suburb the line names. Measured on public registers and sample statements, not yet on lenders’ statements. Confidence 99.0% in the category. 106 of the 106 payee names the register of places read landed on an acceptable category, on 5 November 2026; lenders’ statements have not yet measured it. Given under the balanced setting. As given LUNE CROISSANTERIE FITZROY VIC AUS Card xx7113 Normalised LUNE CROISSANTERIE FITZROY VIC AUS CARD Words matched nothing matched Direction money out Rule no rule matched Register Lune Croissanterie, postcode 3065, from Register of places; 2 places by that name, all on this line

What it does not know

3 payees out of 19 rows, biggest first. Every one of them says unknown rather than other, because a label that hid the gap would leave nobody able to tell a payee this read from a payee it invented a line for. Adding one is a change to a list rather than a change to the software.

1 of these was recognised and deliberately not categorised — a payment to a bank does not say whether it was a mortgage, a card, a fee or the person moving their own money. They are marked, so that nobody is sent to identify a payee somebody has already identified.

Payee, as the feed wrote it	Times	Total	Which
WOOLWORTHS TEAM BANK	1	$400.00	recognised as `WOOLWORTHS TEAM BANK`, and deliberately not categorised
MANNIX & CO 8821	1	$310.00	nothing matched — this is one to add to the rules
NITHIKS KITCHEN ROZELLE 4421 AUS	1	$63.00	nothing matched — this is one to add to the rules

What went where

Category	Group	Rows	Money	Third party named
Salary or wages	income	1	$2,320	0.0%
Rent	essential	1	$2,150	0.0%
Other income	income	1	$1,800	0.0%
Transfer	transfer	2	$1,160	0.0%
unknown	unknown	3	$773	33.0%
Utilities	essential	1	$186	100.0%
Groceries	essential	1	$143	100.0%
Gambling	risk	1	$120	100.0%
Transport and fuel	essential	1	$88	100.0%
Phone and internet	essential	1	$70	100.0%
Eating out and takeaway	discretionary	3	$41	67.0%
Buy now pay later	commitment	1	$37	100.0%
Shopping	discretionary	1	$33	0.0%
Health	essential	1	$25	0.0%

Money is the absolute value of every row in the category, so a credit counts as money moved rather than as a negative expense. The category groups are the ones a serviceability calculation distinguishes between, which is why there is no line for restaurants separate from takeaway.

The transactions

One a line: a description and an amount, separated by a comma or a tab. Credits positive, debits negative. A CSV a bank exported reads too — the layout is worked out from the data. Up to 20,000 rows in one go.

Or a file

A PDF statement, a CSV, an Excel workbook, or an OFX or QIF file. Four megabytes, which is a great many years of transactions. It is read in this request and kept nowhere.

Setting

Strict: The rules alone. Every answer is a name or a kind of business a rule read. Balanced: The rules, the bank’s merchant code where a connection sent one, a register of Australian places, and a modelled reading where it keeps its category at 95% precision or better. Broad: More answers. The bank’s merchant code where a connection sent one, a register of Australian places, and a modelled reading where it keeps its category at 90% precision or better. Complete: Every line gets a category. Read as broad reads, and where nothing read a line, the model’s best guess, marked as one: money out as a spending line, money in as a transfer. Never income, a commitment or gambling on a guess, and not held to a precision floor.

Show me the mixed example Show me a whole invented statement

Nothing is kept: what you paste is read, then gone

What happens to what you paste. It is read while the request is open and then it is gone. Nothing is written to disk, nothing is written to a log, nothing is put in a database — there is no database here to put it in. The transactions travel in the request body rather than the address bar, so this reading has no link, which means there is no link to leak.

The cost of that is honest too. Reload and the reading is gone, and the form re-sends what you gave it from your own browser, because there is nothing on this end to refer back to.

No account, no email address and no tracking cookie; visits are counted without one. Nothing here calls a hosted model: the rules are a text file in this repository and the whole prediction happens in this process.