## What
Auto-categorizes transactions with AI (Gemini) for **pro +
AI-consented** users when no automation rule already matched. Ships the
full backend **behind a Pennant flag, off by default**, so it's
mergeable and testable in isolation; the UI is a deliberate follow-up.
## Why / cost
Prod check first: ~20k txns/month, **~52% of pro-user transactions are
uncategorized** after rules. At Gemini Flash-Lite rates the cost is a
**rounding error** — ~$0.13–$0.75/month for all pro users, single-digit
dollars even on full Flash. So the model is chosen for accuracy, not
price; the real constraints are trust, accuracy and privacy.
## How it works
**Two tiers** (every transaction is covered, rules are an optimization
on top):
- **Tier 1 – label** — a queued listener runs *after* the synchronous
rules; if still uncategorized and the user is eligible, the model picks
a **leaf** category (referenced by numeric index, never a UUID, so it
can't hallucinate one). Auto-applied only above the **label bar**
(`0.7`); below → left blank, no nag. Tagged `category_source = ai` +
`ai_confidence`, fully reversible.
- **Tier 2 – learn** — above the higher **rule bar** (`0.85`) *and* a
clean merchant key *and* the model flags the merchant unambiguous → the
merchant is appended to a single **ai-owned** automation rule for that
category (OR'd conditions, not rule-sprawl), so future transactions
match for free and consistently. AI rules sit at the lowest priority;
**user-owned rules are never touched**.
**Self-heal + signal** — when a user overrides an AI category, a
`category_correction` is logged (calibration signal, bucketable by
confidence) and the offending merchant condition is dropped from the ai
rule (deleted if empty). User rules and manual categories are untouched.
**Safety** — config kill switch + pro + active consent + gradual Pennant
rollout. Dedicated `ai` queue so Gemini never blocks bank syncs.
Encrypted (client-side) transactions are never sent.
**Backfill** — `ai:categorize-backfill {user}`, explicit opt-in,
batched, learns rules as it goes.
## Data model
- `transactions`: `category_source`, `ai_confidence`,
`categorized_by_rule_id`
- `automation_rules`: `origin` (`user`/`ai`)
- new `category_corrections` table
## Screenshots
<img width="921" height="384" alt="image"
src="https://github.com/user-attachments/assets/f04c2a03-b39e-4a3d-81eb-ecf26eaefb83"
/>
## Summary
- add creditor/debtor fields to transactions with raw_data backfill
- store counterparties on bank sync and CSV/XLS imports
- add creditor/debtor filters plus hidden table columns
## Tests
- php artisan test --compact tests/Feature/LocalizationTest.php
tests/Feature/OpenBanking/TransactionSyncServiceTest.php
tests/Feature/TransactionFilterTest.php
- npm test -- resources/js/lib/file-parser.test.ts --run
- vendor/bin/pint --dirty --format agent
Note: `npm run types` still has pre-existing unrelated errors; no
creditor/debtor/import-related errors remained in filtered output.
## Screenshot
<img width="1241" height="676" alt="8odYWtFcvUM"
src="https://github.com/user-attachments/assets/55653485-d588-4beb-9e6a-5c7c81ba7cf8"
/>
## Summary
- **Server-side filtering/sorting/search**: With encryption removed, all
transaction filtering, sorting, and full-text search (LIKE on
description + notes) now happens via SQL queries instead of client-side
IndexedDB with decryption
- **Cursor pagination**: Transactions are returned with cursor-based
pagination (default 50 per page, "Load More" pattern) instead of loading
all into IndexedDB
- **Removed IndexedDB dependency**: The main transactions page no longer
depends on Dexie/IndexedDB for data or sync — other pages (categorize,
accounts/show, budgets/show) are unaffected and will be migrated
separately
## Changes
| File | Change |
|------|--------|
| `app/Models/Transaction.php` | Added `scopeApplyFilters` query scope |
| `app/Http/Requests/IndexTransactionRequest.php` | New form request for
filter validation |
| `app/Http/Controllers/TransactionController.php` | Server-filtered
cursor-paginated index, refactored bulkUpdate to reuse scope |
| `resources/js/pages/transactions/index.tsx` | Major rewrite —
server-driven filtering, sorting, pagination via `router.visit()` |
| `resources/js/components/transactions/transaction-filters.tsx` |
Search always enabled (removed encryption key dependency) |
| `resources/js/types/transaction.ts` | Added `ServerTransaction` type |
| `resources/js/contexts/sync-context.tsx` | Simplified — removed
transaction sync service |
| `tests/Feature/TransactionFilterTest.php` | 20 new Pest tests for all
filter scenarios |
## Test plan
- [x] Run `php artisan test` — all 523+ tests pass
- [x] Navigate to `/transactions` — paginated results load from server
- [x] Apply each filter type (date, amount, category, account, label,
search) and verify server round-trip
- [x] Test "uncategorized" category filter
- [x] Test combined filters
- [x] Test sorting by date, amount, description (ascending/descending)
- [x] Test "Load More" button for cursor pagination
- [x] Test bulk operations (select some → update, select all filtered →
update)
- [x] Verify URL persistence (refresh page with filters in URL)
- [x] Test search on both description and notes fields
Víctor Falcón