Confidente

A Rails + PWA app that treats food sensitivity identification as a structured experiment. Generates scientifically-designed meal plans, logs meals and symptoms, controls for confounders, and uses statistical modeling to surface ingredient-symptom correlations with actual confidence. The hardware that feeds it lives in Citizen Science.

App

App/ — The Rails + PWA application

• Overview — App purpose and architecture
• App/Specs/ — Feature specifications (hypothesis engine, symptom correlator, meal plan generator, daily control quality score)
• App/Science/ — Statistical methodology (ANOVA, experimental design, confounder control)
• App/Architecture/ — Schema, tech stack, constraint solver research
• App/Foods/ — SIGHI seed data by food category

Reference Data

Mast Cell Reference/ — SIGHI food compatibility data and medication manual summaries. Primary source for the app’s food seed data.

• SIGHI Medication Manual Summary
• Source document index and version notes

Maturity

Stage: Pre-alpha, active development. The app is the only component in active development. The statistical methodology is designed but untested on real data. The food seed data (SIGHI) is imported; the hypothesis engine and symptom correlator are specified but not built. The hardware that was originally part of this project lives in Citizen Science.