What this article is about
TaxItEasy doesn't just match bank transactions to invoices — it pays attention to what you do with those matches. Confirm a suggestion, reject one, unlink a match, or pick a category, and the system remembers. This article explains exactly what is learned, what the learned memory is allowed to do (and, more importantly, what it is never allowed to do), and where your data stays.
One thing up front: learning starts from your live usage. It builds as you confirm and correct — it is not pre-trained on anything.
Trusted vendors
The core of the memory is a per-company list of trusted vendors. A vendor becomes trusted after you've confirmed 3 matches for it, as long as you haven't been rejecting its suggestions too often (a high reject rate keeps a vendor out of the trusted list, and frequent rejections scale the effect back down).
Once a vendor is trusted, its future match candidates get a confidence boost of up to 12 points on the 100-point match score. In practice: a plausible-but-not-obvious payment from a vendor you've confirmed many times ranks higher and surfaces sooner than the same payment from a vendor the system has never seen you confirm.
Confirming a match also teaches the system new spellings: the counterparty name as your bank wrote it is saved as an alias for that vendor, so the counterparty signal recognises the vendor next time even when the bank mangles the name.
The guardrail: learning never silently escalates
This is the most important property of the whole system. The learned boost never pushes a score across an automation boundary from below:
- A score below 70 can never reach the auto-link tier through learning alone.
- A score below 90 can never become a silent auto-match through learning alone.
- A match that is already strong gets no boost at all — there's nothing to improve.
In other words: learning re-ranks and proposes. It never manufactures a link, and a suggestion cannot quietly turn into a silent match just because you've confirmed that vendor before. Anything that crosses into automatic territory got there on the strength of the actual evidence — invoice number, amount, date, counterparty — not on reputation.
Category learning
The memory also covers categories, in two places:
- Invoice categories. When you pick the same category for the same vendor 3 times, the memory activates and that category is suggested for the vendor going forward.
- Transaction categories. Categorising a bank transaction directly — one without a receipt attached — teaches the same memory. This works for owners and for tax advisors, including bulk actions.
Suggestions are always suggest-only. They appear as a chip — "Suggested: {Label} · Accept" — above the category field; nothing is ever booked automatically. If a vendor genuinely varies (the system has seen 3 or more different categories for it), it stops suggesting for that vendor instead of guessing. A single correction makes your newest choice win and restarts the count.
Timing note: learning takes effect immediately, but the suggestion chip appears after the next processing run (a bank-statement import or the nightly sweep) — not in the same second you taught it.
Precedence: Rules > Learned > AI
When the system decides what to suggest for a transaction, three sources are consulted in strict order:
- Your rules — an explicit matching rule you've defined always wins.
- Learned memory — your company's confirmed history comes next.
- AI suggestion — only for the long tail neither of the above covers, and low-confidence AI suggestions are dropped rather than shown.
You are always the top of the hierarchy: anything you've stated explicitly outranks anything the system inferred.
Everything stays in your company
- Learning is strictly per company. There is no cross-customer pooling — what other TaxItEasy users confirm has zero effect on your matches, and yours has zero effect on theirs.
- It's deterministic counting, not model training. No external service sees your confirmations.
- Vendor names that aren't linked to a vendor record are stored only as irreversible cryptographic fingerprints, never as readable text.
- When you delete your account, the learned data is purged along with everything else — what the system learned from you is deleted with you.
Related
- Why this matched — and how to undo it — the explanation behind every match
- How the matching pipeline works — the four signals and the tiers
- Creating a custom matching rule — the explicit layer that outranks learning
- Booking transactions for your clients — advisor actions teach the same memory
- Delete your account and export your data — what erasure covers