For reviewers
MyFoodFit is an early-stage, research-aware, rule-based product designed to help people interpret food labels. The system is in active beta. Outputs are informational and non-diagnostic.
What is implemented (v6.1.0)
- Deterministic 5-layer scoring engine (macronutrients, NOVA, allergens, micronutrients, condition-specific)
- 35+ dietary profiles covering allergens, medical conditions, life stage, and medication
- Barcode scanning with 7-tier lookup chain; offline bundle of 17,845 UK products
- AI photo scanning via GPT-4o Vision (meals and nutrition labels)
- Natural language food discovery and constraint-aware recipe search
- Explainable signals tied to documented rules, with confidence indicators where data is incomplete
- Scoring validation partnership with Solent University underway
What this is not
- Not a medical device
- Not diagnostic or therapeutic
- Not predictive
- Not clinically validated
- Not a substitute for professional advice
How to evaluate this product
Explainability
Can outputs be traced to rules?
Transparency
Are sources and limitations visible?
Appropriateness of scope
Are claims restrained?
Handling of uncertainty
Is ambiguity surfaced conservatively?
Supporting material
Current status
The product is at v6.1.0 (Build 47 on TestFlight) with an active Solent University validation partnership. Rules and thresholds continue to be refined. Evaluation and critique are welcome. Review does not imply endorsement. Contact is via the existing contact page.