Shelf Scan
Point your camera at a grocery shelf. Get nutrition info on every product.
Drop an image here or tap to upload
JPEG, PNG, WebP, or iPhone photo · Works best with a clear shot of a shelf
Analyzing up to 25 products per scan ·
How it works
- 1Upload a photoSnap a grocery shelf or pick an image from your camera roll.
- 2Every product is foundDetection locates each item on the shelf, not just the one in front.
- 3Each one is identifiedBrand, product and variant are read off the packaging.
- 4See the resultsNutrition info for everyone — a fit score if you’re signed in.
What you should know
Privacy
We don’t blur faces yet, so try to avoid people in frame. We never store your images.
Signed in vs. not
Not signed in, you still get full nutrition facts, processing level, and flagged ingredients. Sign in for a fit score against your allergies and goals.
Scan settings
Change how many products are analyzed per scan any time in Settings.
Claim Check
Photograph the front of one product. See if its claims match what's actually inside.
Drop a product photo here or tap to upload
Front of package only · JPEG, PNG, WebP, or iPhone photo
One product per check · Scanning a whole shelf?
How it works
- 1Photograph a productGet a clear shot of the front of the package.
- 2We read its claimsMarketing claims and label text are extracted.
- 3We check the factsNutrition data is compared against what’s claimed.
- 4See the verdictA clear read on which claims hold up — and which don’t.
What you should know
Privacy
We don’t blur faces yet, so try to avoid people in frame. We never store your images.
One product at a time
Claim Check works on a single product’s front-of-package claims against its real nutrition facts. Scanning a whole shelf? Use Scan instead.
AI-generated checks
Claim-checking is generated by AI and may miss nuance — always check the full nutrition label yourself for the final word.
What This Detects
False "Natural" claims
Products marketed as natural but containing synthetic additives
Hidden sugar patterns
Foods labeled "healthy" with statistically anomalous sugar content
Ingredient misrepresentation
Marketing that obscures ultra-processed ingredients
Regulatory gaps
Claims that exploit loose FDA labeling regulations
Ingredient Analytics
Search any ingredient for its safety profile, regulatory status, and what it's commonly found with.
Popular searches
How it works
- 1Search an ingredientType a name, or tap one from the directory below.
- 2We pull the recordsIARC, EFSA/FDA, CSPI, NOVA and ADI data for that ingredient.
- 3A composite scoreThree weighted tiers combine into one concern score, shown with its working.
- 4See what it travels withThe ingredients, categories and products it most often appears alongside.
What you should know
Where the data comes from
IARC Monographs (WHO), EFSA/FDA regulatory records, the EU "Southampton Six", JECFA/EFSA ADI limits, CSPI Chemical Cuisine, and the NOVA classification.
Scores are not verdicts
A composite score summarises what authoritative bodies have published — it is not a medical judgement, and an ingredient with limited data scores as unknown rather than safe.
Co-occurrence is observational
What an ingredient commonly appears with is drawn from USDA product records. It describes the market, not causation.
Ingredient Directory
354 ingredients with authoritative safety data (IARC, EFSA/FDA, CSPI, NOVA, ADI) — click any to analyze
Showing 354 ingredients
Sources: IARC Monographs (WHO), EFSA/FDA regulatory records, EU “Southampton Six”, JECFA/EFSA ADI limits, CSPI Chemical Cuisine, and the NOVA classification. Dot colour = composite concern level.
Know what's on the shelf
before it's in your cart
Nutritionell turns a single photo of a grocery shelf into a clear, personalized breakdown — identifying every product, analyzing its ingredients and nutrition, and scoring it against your dietary profile.
Our Mission
The modern grocery store is designed for brands, not shoppers. Consumers are left to decode flashy marketing and obscure ingredient lists on their own, one product at a time. Nutritionell exists to put scale and simplicity at the heart of nutrition intelligence — so anyone can walk down a grocery aisle and instantly understand what they're actually buying, without becoming a food-label expert first.
The problem — and our solution
Where grocery shopping stands today, and what Nutritionell does about it.
The Problem Space
- Shoppers decide amid flashy marketing and obscure chemical ingredients.
- Health-conscious consumers are left to research brands and ingredients themselves.
- Current apps create information bottlenecks — slow, item-by-item scanning.
The Solution
Nutritionell
From a single grocery-shelf photo, identify each unique product, run nutrition analysis, and get user-specific recommendations — replacing singular product lookup and hours of research.
How it works
Three steps from shelf to smarter choices.
What you can do
Everything you need to shop with confidence.
Why Nutritionell
What sets this apart from a barcode scanner.
⚠️ A Note on Medical Advice
Nutritionell is an informational tool designed to help you understand product ingredients and nutrition more easily. It is not a medical device and does not provide medical advice, diagnosis, or treatment. Scores and recommendations are generated by AI models and general nutrition data, and may be incomplete or inaccurate. Always consult a qualified healthcare provider or registered dietitian for guidance on your specific health conditions, allergies, or dietary needs before making decisions based on this app.
Contact Us
Nutritionell is built by a small team who believe grocery shopping should be transparent, not overwhelming. Reach out any time.
Meet the Team
Steve Lanciotti
The Alchemist
Computer vision and vision-language models for accurate product detection, identification, and analysis.
Priyanka Banerjee
The Vanguard
User-facing application for instantaneous nutrition insights, field testing, and milestone delivery.
Yu-Sheng Lee
The Architect
Cloud infrastructure, database architecture, and backend pipelines for real-time processing.
Najmeh Rahimi
The Oracle
Subject-matter expertise, domain logic, and user feedback loops for AI vision reasoning.
Get in Touch
Questions, feedback, or partnership ideas?
We'd love to hear from you — whether it's a bug report, a feature request, or just general feedback on how Nutritionell is working for you.
nutritionell@gmail.comSettings
Preferences that control how Nutritionell looks and scans your shelf. These are saved on this device.
Appearance
AvocadoChoose a color scheme. Light themes are the default; dark themes are available too.
Light
Dark
Detection model (YOLO)
YOLO11nWhich detector finds the products in your shelf photo. Lighter models scan faster; heavier ones can catch more small or tightly-packed products. Your choice is used for the YOLO step of every scan.
Applies to the Scan tab. Default is YOLO11n.
Maximum products per scan
25How many products a single shelf photo will identify and score. A lower number makes each scan noticeably faster and more reliable; a higher number captures more of a busy shelf but takes longer and is more likely to time out.
Applies to the Scan tab. Default is 25.
Checking your session…
Sign in to continue
Profile, Goals and My Plan need an account. Everything else works without one.
Sign in to continue
Profile, Goals and My Plan need an account. Everything else works without one.
Sign in to continue
Profile, Goals and My Plan need an account. Everything else works without one.