Jev in production › Document and record classification

Penny

Jev suggests storage location and product group for scanned grocery receipts before import into Grocy; low-confidence suggestions are left unassigned.

Open on GitHub ↗

The source publishes no measured number.
Use
Document and record classification
Industry
Productivity
Form
Open-source tool
Stage
In production
Plugs into
Grocy
Listed
2026-10-25
Found via
github
Repository
YuryFilipovich/Penny
Stars
8
Forks
0
Last push
2026-10-27
Language
TypeScript
License
Apache-2.0
Penny screenshot
docs/images/Penny1.png in the YuryFilipovich/Penny README, Apache-2.0; shown from GitHub.

The README opens with

Penny turns grocery receipt PDFs into a reviewable product list, then imports confirmed purchases into Grocy. It uses a Svelte 5 + TypeScript frontend and a Node.js backend. OpenAI extracts receipt lines; Jev suggests storage locations and product groups. A person reviews every field before anything is written to Grocy.

1. Upload a PDF receipt (up to 20 MB and 30 pages), or open the built-in sample without using an API key. 2. Penny sends the PDF to OpenAI for product extraction. It shows the original alongside editable names, quantities, units, line prices, store, date, and currency. Searchable PDFs can highlight the source line for a product. 3. Jev suggests a storage location and product group. Low-confidence suggestions stay unassigned....

Badge

For the project's own README, linking back here:

Listed in Jev in production

Also used for document and record classification