Category Taxonomy
Every transaction in Expense Radar AI is classified into one of 26 fixed categories. These categories are organised into three buckets that reflect how money flows through your finances.
How your money is classified
Income
5 categories for money coming in
Savings & Investments
2 categories for money set aside
Expenses
19 categories for money going out
Income(5)
Income
Salary, wages, freelance invoices paid, and other earned income
Tax Return
Tax refunds received from the tax authority (e.g. Belastingdienst)
Allowances
Regular government or employer allowances received (e.g. toeslagen)
Money Gifts
Money received as a gift from family or friends (birthdays, holidays, occasions)
Investment Income
Return on investments: dividends, interest, coupons, and realised capital gains
Savings & Investments(2)
Savings
Transfers to savings accounts
Investments
Transfers to investment/brokerage accounts (NOT dividends or returns)
Expenses(19)
Housing
Rent, mortgage, repairs, furniture, home services
Groceries
Supermarket and food-store shopping for items consumed at home
Dining Out
Restaurants, cafés, bars, takeaway, food delivery, lunch spots
Utilities
Electricity, gas, water, internet, mobile phone
Transport
Fuel, public transit, taxi/rideshare, car costs and maintenance
Insurance
All insurance premiums (health, home, car, life)
Taxes
Tax payments made to the authorities (income tax, property tax, fines)
Health
Medical appointments, pharmacy, gym and fitness
Personal Care
Haircuts, toiletries, cosmetics, beauty services
Shopping
Clothes, electronics, general retail and online shopping
Subscriptions
Streaming, software, and other recurring digital subscriptions
Travel
Travel, holidays, flights, hotels and accommodation
Entertainment
Events, cinema, concerts, hobbies, leisure
Education
Courses, books, training, tuition
Childcare
Nursery, babysitting, school fees, kids activities
Pets
Pet food, vet visits, pet supplies and services
Gifts
Gifts purchased for other people (NOT charitable donations)
Charity
Donations to charities and non-profit organisations
Other
Catch-all for transactions that do not fit any other category
Why fixed categories?
Expense Radar AI uses a fixed set of categories rather than letting you create your own. This is a deliberate design choice with several benefits:
Consistent tracking
Fixed categories mean your spending data is comparable month over month and year over year. You can spot trends reliably because categories never shift or overlap.
Smarter AI
The AI learns from a well-defined set of categories, making its automatic assignments more accurate. A consistent taxonomy means fewer corrections over time.
Simpler budgets
Budgets map directly to categories. No ambiguity about where a transaction belongs, no duplicate categories to reconcile, and reliable reporting out of the box.
The trade-off: you cannot create custom or user-defined categories. We believe the 26 categories above cover the vast majority of personal finance needs. If you find a transaction that genuinely does not fit, the “Other” category is available as a flexible catch-all.
How AI categorises your transactions
Expense Radar AI automatically assigns a category to every transaction that arrives from your bank. The process is designed to be fast, accurate, and to improve over time based on your corrections.
Instant recognition
When a new transaction arrives, the system first checks whether it has seen the same merchant before. If so, it instantly applies the same category — no waiting required.
Merchant identification
For new transactions, the system reads the bank's raw description to identify the merchant. It then checks a merchant dictionary — if you have ever corrected a category for this store or service, that correction is used automatically.
AI classification
Only when the system encounters a genuinely new merchant does it call the AI. The AI receives the transaction description and the full category table, then assigns the best-fitting category. The result is remembered for future instant matching.
Your corrections improve accuracy
When you change a transaction's category, the correction propagates throughout the system. The merchant dictionary is updated, the cache is refreshed, and all future transactions from that same merchant use your corrected category. Your past corrections are also provided to the AI as context, making it smarter over time.
Graceful fallback
If the AI is temporarily unavailable, transactions receive “Other” as a fallback category rather than failing. You can recategorise them later at any time.