MyFoodFit

Grant appendix PDF export

This view is designed for PDF export and grant attachments. It combines reviewer materials, grant appendix, and evidence log in a single print-ready document.

For best results, use Chrome or Safari and select 'Save as PDF'.

Direct link for reviewers

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

  • Rule-based food evaluation logic
  • Explainable signals tied to documented rules
  • Preference-driven evaluation
  • Confidence indicators where data is incomplete
  • Active beta usage

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 early-stage and evolving. Evaluation and critique are welcome. Review does not imply endorsement. Contact is via the existing contact page.

Grant appendix: scope and evidence checklist

This checklist summarises current scope and references observable functionality only. It is intended to accompany formal review and does not introduce new claims or language not present elsewhere in the documentation.

Product status

  • Early-stage product
  • Active beta
  • Research-aware
  • Rule-based logic
  • Not a medical device

What is implemented

  • Explainable food evaluation
  • Preference-driven scoring
  • Confidence indicators for incomplete data
  • Transparent rule descriptions
  • User-facing explanations

What is not claimed

  • No diagnosis
  • No clinical validation
  • No health outcome prediction
  • No personalised treatment guidance

Supporting documentation

Evidence log

This log records observable product capabilities as they are implemented. Entries describe implementation, not outcomes. Inclusion does not imply validation or endorsement.

Current release (v6.1.0)

March 2026

Status: Build 47 on TestFlight

Implemented and observable

  • Personalised RAG scoring (0–100, Red/Amber/Green) across 35+ dietary profiles
  • Multi-layer scoring covering nutrition, processing level, allergens, micronutrients, and condition-specific thresholds
  • Barcode scanning with multi-source data resolution
  • AI photo scanning for meals and nutrition labels (independently benchmarked and validated)
  • Find Food: AI-powered natural language food discovery
  • Find Meals: constraint-aware recipe search
  • Smart swap suggestions (23 categories, 100+ alternatives)
  • Personal Cookbook for saving and re-cooking scored recipes
  • Offline-first scanning bundle for popular UK products
  • Weekly progress tracking with personalised insights
  • All 14 EU allergen detection
  • Profile-weighted micronutrients (pregnancy prioritises folate, menopause prioritises calcium, ADHD prioritises iron)
  • Whole food protection, ultra-processed food detection, and processed meat flagging
  • Partial data scoring: products scored even with incomplete nutrition data
  • Comprehensive global food database with extensive UK product coverage
  • Comprehensive automated test coverage across all scoring and integration paths
  • Scoring validation partnership with Solent University

Explicitly not claimed

  • No medical diagnosis
  • No personalised medical advice
  • No predictive health outcomes
  • No clinical validation

Earlier versions

Earlier versions (v2.x and below) focused on foundational explainability and basic preference-based scoring. Coverage, scoring depth, and feature set have evolved substantially through v6.1.0. Prior versions are not presented as current evidence.

How this log should be used

  • As a scope reference for reviewers
  • To support transparency and auditability
  • To prevent misinterpretation of beta functionality