Features
Pre-launchFourteen capabilities, in four families: understand, watch, ask, control.
01Understand
The original label stays intact. Beside it, what Expaly makes of it: a merchant, a category, an explanation — and the confidence that goes with it.
Current accounts and cards land in a single normalised model, with the source of every line. Filter by account, category, period and status; search covers both the original label and the reading.
One list to read instead of three apps to open.
Every operation has a two-column record: the original bank data, never altered, and the product's reading, clearly marked as a reading.
You can see where each statement comes from, and what is interpretation rather than fact.
Behind an abbreviated, truncated or Hebrew label, Expaly names the real merchant. When it cannot, it writes "merchant not identified" rather than guessing.
You know who charged you, not just how much.
Operations are filed in a stable taxonomy: groceries, restaurants and cafés, streaming, bank fees, cash withdrawals, and the rest. What isn't classified stays visibly unclassified.
Totals per category, without filing every line yourself.
A Hebrew label stays on screen exactly as it is, in monospace, in its own writing direction — and gets an explanation next to it, in French, English or Hebrew.
You no longer need to read financial Hebrew to understand your statement.
Example on synthetic data.
02Watch
Expaly looks at your series rather than your lines. A recurring charge that rises, a merchant never seen, an unusual rhythm: each one dated, quantified and tied back to the operation it came from.
Attention
13 August 2026Netflix
monthly · next on 6 Sept 2026
Old amount
49.90 ₪New amount
54.90 ₪+10%What these figures cover
5 June 2026 to 11 August 2026
An alert only holds for the period actually imported. Operations Expaly could not name stay shown as such, never guessed.
Example on synthetic data.
03Ask
No language model over your finances: deterministic engines, on your own operations. The same question always gives the same answer.
Natural-language questions resolved by deterministic engines over your data: spending across a period, recurring payments, subscriptions on the rise, unusual operations, bank fees. No language model, no financial data sent to an external service.
A numeric, reproducible answer you can verify line by line.
A summary per period: total, comparison against the previous period of equal length, breakdown by category — and, at the top, what the numbers actually cover and up to which date.
A figure that also tells you what it doesn't say.
What these figures cover : 5 June 2026 to 11 August 2026
Example on synthetic data.
04Control
Every reading Expaly makes is a proposal. You confirm it, you correct it, and the product remembers. The household decides who sees what.
Any proposed reading can be confirmed or corrected. A "to confirm" queue gathers what is waiting on you, and Expaly remembers your corrections for next time.
The product learns your actual life: your plumber, your school, your building manager.
The household is the isolation unit for data: two adults included, each with their own access and role. An extra member joins for 10 ₪ per month.
A shared view without a shared account or a swapped password.
What isn't open yet is the automatic bank connection. Sign up to be told when it is.
