Transaction enrichment and categorization
Turn raw transactions into categorized data you can build on
Finory identifies merchants, sorts transactions into 78 categories and extracts the people and businesses behind them, trained on Malaysian banking data.
Raw bank data
Real transactions processed by Finory. Reference numbers shortened.
2M+
Unique transactions our models are trained on
95%
Average accuracy on categorization and merchant identification
Every day
Powering our own consumer apps
What you get
Every transaction, explained
Merchant identification
Clean merchant names from cryptic bank descriptions, with 352 brands recognized.
78 categories
From salary and rent to food delivery and utilities.
Entity extraction
Who is behind each payment: an employer, a business or a person.
Income and behaviour signals
Salary detection, rent, utilities, investments and loan disbursements.
Built for Malaysia
Global models don’t speak Malaysian banking
Malaysian statements are full of local payment rails, eWallets and merchants that generic enrichment tools misread. Our models are trained on Malaysian bank, credit card and eWallet data, so they understand them.
Use cases
What banks build with enriched data
Spending insights in your app
Add personal finance features, such as spending breakdowns and budgets, that keep customers coming back.
Faster credit decisions
Verified income, commitments and behaviour, ready for underwriting.
Open finance products
Turn shared customer data into products, offers and insights.
Fraud and risk signals
Spot window dressing and round-tripping. It’s the same engine behind our document fraud detection.
Preparing for open finance in Malaysia?
When customer data starts flowing between institutions, raw transactions won’t be enough. Enriched data is what turns access into useful products.
Transaction enrichment and categorization: frequently asked questions
What is transaction enrichment?
Transaction enrichment turns cryptic bank transaction descriptions into clean, structured data. Finory identifies the merchant or brand behind each transaction, assigns a category and extracts the business or person involved, so the data can power credit decisions, insights and personal finance features.
How does Finory categorize Malaysian bank transactions?
Finory sorts transactions into 78 categories and recognizes 352 brands, using machine learning models trained on more than 2 million Malaysian transactions, with 95% average accuracy across categorization and merchant identification. It understands local payment rails and eWallets such as DuitNow, Touch ’n Go eWallet and ShopeePay.
What data can Finory enrich?
Finory enriches transactions from bank statements, credit card statements and eWallet statements from all Malaysian banks and providers, either from uploaded documents or from raw transaction data sent through its API.
Can Finory detect recurring payments and instalments?
Yes. Finory detects recurring payments, instalment months and how often a customer buys from each brand, in addition to categorizing every transaction.
How does transaction enrichment help banks prepare for open finance?
Open finance gives banks access to customers’ transaction data, but that data arrives raw. Finory enriches it with merchants, categories and entities, so banks can build lending, insight and personal finance products on it. Banks can start today with statements and switch to shared data later.
See your transactions enriched
Share a sample of your transaction data, and we’ll show you what Finory returns.
Book a demo