Plattr

Know what's behind every bite.

Live app: plattr.replit.app · one-click demo accounts at /demo-account

Scan a food's QR code or barcode and read its product passport: farm origin, soil, water, feed, welfare, animal health, certifications, safety testing, a hazard outlook for the region, and nutrition. Every passport gets a transparent 0-100 Plattr score (A+ to F), and every fact carries a text label saying how well it is backed up, so a producer's own claim never passes as a verified record.

Plattr is a student prototype. The product passports in the demo are fictional SAMPLE data.

For shoppers

• Find a food (/explore): search a name, brand or barcode, scan with the camera, or upload a photo. Results mix Plattr passports with live Open Food Facts records, labelled Community record.
• Read the passport (/food/:id): every fact shows an evidence chip - Verified record, Document on file, Producer-declared or Not provided - plus the score broken down section by section, a hazard outlook, nutrition facts and personal notes.
• Profile (/profile): allergens, health conditions, dietary preferences, optional lab values and today's intake. Warnings are deterministic, on-device and informational only, never medical advice.
• Grocery trip (/shop) and My groceries (/history): build a cart, see its running score and grade, finish the trip and keep a history with an overall score per timeframe.
• Map (/map): the farms behind what you bought, with distances from your location (which never leaves the browser).
• Health trends (/trends): logged intake charted over 7 to 90 days against FDA Daily Values. Unlogged days are gaps, never zeros.
• USDA plant lookup (/plant-lookup, /est/:number): the real federal record behind the establishment number on a meat pack.

Find a food	Product passport

Grocery trip	My groceries

Health trends	USDA plant lookup

For producers

• My farm (/producer): location, acreage, herd, water tests, fertilizers and pesticides, health records and a parasite watch list. Land farms, fish farms and wild-catch fisheries are all supported.
• Product passports (/producer/passports): fill in the sections, watch the score update live, and get a QR code that opens the passport. Producer entries show as Producer-declared or Document on file and earn half the points of a verified record. Nothing is uploaded or inspected in this prototype.

My farm	Passport builder

How the score works

The score measures how much of a food's story is documented and backed up. It is not a food-safety or health guarantee. One pure function, src/passport/score.ts, does the work and every part is shown on screen.

Evidence chip	Weight
Verified record	1.0
Document on file	0.9
Producer-declared / Community record	0.5
Not provided	0

Each section earns its average weight times its maximum points (farm origin 20; feed, welfare and animal health 15 each for animal products; soil 25 and water 20 for plant products; certifications and safety testing fill the rest to 100). Grades: A+ 97-100 · A 90-96 · B 80-89 · C 70-79 · D 60-69 · F below 60. The full rubric is on /how-it-works.

Real vs sample

• Real: USDA FSIS plant directory, recalls, sampling and humane-handling data behind the plant lookup (snapshots in public/data); regulation quotes behind label claims; FDA Daily Values and the cited reference values in src/health/references.ts.
• Sample: the nine passports, their farms, the rancher profiles and the two demo accounts are fictional and marked SAMPLE on screen. Hazard notes are general background, not findings about any farm.
• Storage: accounts, profiles, carts and producer passports live only in this browser's localStorage. No server, no real auth. Do not enter a real password or real medical data.

The health layer's design reasoning is in docs/HEALTH-ARCHITECTURE.md.

Run it

npm install
npm run dev      # Vite dev server
npm test         # Vitest
npm run build    # type-check, then production build into dist/
npm run start    # serve the production build on $PORT or 4173

Needs a current Node.js LTS. No environment variables or secrets. To deploy on Replit, import the repo and the included .replit file does the rest; for a static host, publish dist/.

Tech stack

Vite + React + TypeScript, react-router, plain CSS · Leaflet with OpenStreetMap tiles · qrcode · @zxing/browser for barcode decoding · Vitest · Vercel static hosting with cookie-free Web Analytics. Everything runs in the browser over static JSON in public/data. No backend and no LLM.

Data sources

USDA FSIS (inspection directory, recalls, sampling, humane handling, labeling guidance) · FDA (food substances, GRAS, color additives, Daily Values) · eCFR and Federal Register · European Commission DG SANTE and EUR-Lex · Open Food Facts (ODbL) · OpenStreetMap contributors.

The research notes, raw source files and the scripts that built public/data live in the team's private working repository. A production PostgreSQL traceability schema is designed but not part of this repo (Figma site).

License

Code under the MIT License. Government data is public domain; Open Food Facts data is ODbL; map tiles are © OpenStreetMap contributors. Sample passports, farms, ranchers and demo accounts are fictional.

Plattr is a student project. It is not legal, medical or dietary advice.

Team

• Parth Dave — @ParthD25
• @ali-findra
• Gabriela Hamdieh — @ghamdieh-create
• Connor Young — @conyoung18
• @Tanvirfs29