When RECEIPT_VISION_MODE=vision (default), uploaded receipt images are sent
directly to the vision-capable LLM (llama3.2-vision via Ollama) instead of
the OCR text excerpt. The model can read logos, stylised fonts, and layouts
that Tesseract OCR mangles (Home Depot, HMSHost/Sergio's, etc.).
Architecture:
- amount + date: always from Tesseract regex (deterministic, never LLM)
- vendor + category: vision LLM when image available, text LLM as fallback
- Fallthrough: if vision call fails for any reason, text path is tried next
- PDF/TXT/HTML receipts: always use text path (not visual media)
Revert instantly without a rebuild:
echo "RECEIPT_VISION_MODE=text" >> /root/odoo/odoo-ai/.env
docker compose up -d agent-service
config.py: add receipt_vision_mode setting (default 'vision')
expenses_agent.py: _VISION_MIMETYPES, _get_vision_mode() helper,
dual-path _parse_receipt_text (b64/mimetype params), _act() passes b64
tests: 92 passing — 4 new vision tests, 2 existing prompt tests
pinned to text mode via _get_vision_mode patch
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- Remove "NeDonald's → McDonald's" from LLM vendor correction examples; the
example was biasing the model to return McDonald's for any ambiguous receipt
(Home Depot, Sergio's/HMSHost). Replace with neutral brand examples and add
an explicit instruction not to substitute a brand name absent from the OCR text.
- Add `net\s*fee` to _TOTAL_RE so MIA Parking kiosk receipts ("net fee: 150.00 USD")
are captured by Pass 1 rather than the max-scan which could pick a larger line.
- Add Step 5b grayscale fallback in receipt_parser: if all binarized PSM attempts
yield < 20 chars, retry OCR on the pre-binarization grayscale image. Fixes
dot-matrix and certain thermal-print fonts destroyed by the 160-threshold.
- Tests: 88 passing (test_net_fee_parking, test_vendor_prompt_does_not_contain_mcdonalds,
test_vendor_prompt_instructs_not_to_guess_absent_brand).
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
LAYAL CAFE ($2.80 instead of $42.90):
- Add (?!\s*tax) lookahead to _TOTAL_RE so "Total Taxes $2.80" is never
confused with the receipt total when OCR drops the "Taxes" word
- Change Pass 1 from matches[-1] to max() so the largest labeled amount
always wins, regardless of line order in the OCR output
United Airlines (Subway/$0/wrong date):
- Add OSD-based rotation correction in receipt_parser.py: after EXIF
transpose, ask Tesseract's orientation-detection engine (--psm 0) what
angle to rotate; applies to receipts photographed lying sideways where
EXIF metadata cannot help
- Add month-name date patterns (DD MON YYYY / MON DD YYYY) to
_extract_date_from_text for airline/hotel receipts that print dates
like "05 MAY 2026" instead of "05/07/26"
85 tests, all passing.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Gas station receipts (Costco, Shell, etc.) print "Total Sale $X.XX" — the
word "Sale" between "Total" and the amount prevented _TOTAL_RE from matching,
causing the Costco receipt to fall through to the max-scan heuristic and
return a garbled OCR value instead of the correct total.
Also add "Net Sale" and "Sale Total" variants for broader coverage.
79 tests, all passing.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- Add _is_likely_bank_statement(): if OCR text has ≥10 lines with dollar
amounts it is almost certainly a bank/card statement screenshot, not a
single receipt. Return skip=True so _act() skips it and adds a note to
the escalations list instead of creating a $1,699 expense line.
- Fix default product selection in _act(): prefer "Meals" over whatever
happens to be first in Odoo's expense product list ("Communication"),
so unrecognised receipts get a sensible fallback category.
- Improve LLM category prompt: remove hardcoded product names (airline →
Transport) that don't exist in every Odoo install; describe business
types semantically so the model picks from the actual available list.
- Mention skipped statements in the final summary message.
- 77 tests, all passing.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- Remove card brands (VISA/MC/Amex) from _SKIP_LINE_RE so card-terminal
lines like "VISA USD$ 36.78" are no longer skipped
- Replace bottom-50% scan with full-text max scan (Pass 2): scans every
line in the receipt and returns the largest dollar amount, correctly
handling display-style receipts that show the charge at the top with
no label (e.g. LAYAL CAFE $40.10 before the item list)
- Update vendor LLM prompt to ask the model to correct OCR garbling
(e.g. "NeDonald's" → "McDonald's") and detect bank statements
- Add 4 new tests covering top-amount, card-terminal, max-beats-items,
and change-exclusion scenarios (71 tests, all passing)
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Receipt quality: replace LLM amount/date extraction with regex.
LLM was hallucinating 2021/2022 dates and returning '198.40 USD' strings.
Amounts now use deterministic regex (Total:/Grand Total:/Amount Due:).
Dates: filename timestamp > OCR regex > today (no LLM date guessing).
LLM only asked for vendor name + product category.
Approval: fix GET /approval/pending 500 by using correct column
name 'started_at' instead of 'created_at' (which does not exist).
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
The llama3.2-vision model was producing unreliable structured data
(wrong vendors, amounts, dates) making expense reports worse than
Tesseract + LLM extraction. Removes _ocr_image_vision(), the
vision JSON fast path in _parse_receipt_text(), _match_category(),
and the vision_ocr_model config setting entirely.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
receipt_parser: change _ocr_image_vision() to extract structured JSON
{vendor,amount,date,time,category} directly from the image instead of
transcribing raw text, so the downstream LLM extraction step is
unnecessary and the two-step error-compounding is eliminated.
expenses_agent: add _match_category() helper to map vision category
labels to expense product names via substring/fuzzy match; add fast
path in _parse_receipt_text() that detects pre-extracted vision JSON
(text starts with '{') and skips the second LLM submit call entirely.
Fix text[:2000] truncation that discarded receipt totals — now keeps
first 1500 + last 1500 chars of long receipts so the grand total at
the bottom is always included.
tests: fix stale test_act_enters_awaiting_confirmation_on_first_pass
(confirmation gate was removed); add TestMatchCategory and three new
tests for the vision JSON fast path and LLM fallthrough.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Adds module-level label and cross-reference to the new doc.
TEST_EXPENSES_AGENT.md documents every test group, case, and the
real-world bug each test guards against (e.g. In-N-Out OCR mismatch).
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
'Restaurant A' vs 'Restaurant Z' differ by 1 char so difflib scores
them at ~91% -- correctly above the 80% threshold. Use clearly
different vendors (Starbucks Coffee vs McDonalds Burger) instead.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
