Zitounix
Health goals, without the daily friction.
AI-powered nutrition analysis that turns a meal photo into structured, confidence-labeled nutrition data — then corroborates every estimate against official food-composition databases before showing a result.
Daily health tracking is fragmented and friction-heavy.
People managing diabetes, weight, or fitness goals juggle several apps — one for meals, one for glucose, one for activity — and manually log every macro by hand. Most AI nutrition tools make the problem worse by returning opaque 'black-box' estimates the user cannot trust or verify.
Zitounix unifies nutrition, activity, connected-device data, goals, and progress into one continuous daily record — and makes the AI layer transparent rather than mysterious.
Generative vision AI, corroborated against ground truth.
A vision-capable LLM analyzes the meal photo and estimates per-component macros, glycemic index, and glycemic load — each tagged with explicit confidence and identity-confidence ratings.
Two-pass visual analysis
Fast scan for quick logging; deep scan runs a parallel independent 'blind visual audit' that cross-checks food identification against the initial pass, plus reasoning mode for ambiguous components.
Confidence-labeled outputs
Every portion, macro, and identification carries a low/medium/high confidence rating with visual evidence and candidate alternatives — the user always sees how certain the AI is, and why.
Corroboration against 11 official databases
AI estimates are paired with independent lookups from CIQUAL/ANSES, USDA FoodData Central, NEVO, and 8 more regional sources — the user sees the AI estimate next to the official calculation and any meaningful disagreement.
AI period analysis & daily plans
Generative AI summarizes meals, activity, and metrics into a personalized narrative over N days, and proposes structured daily plans adapted to the user's objective (glucose, weight, fitness).
AI output is never accepted blindly. Every estimate is corroborated against official data before it reaches the user. Uncertainty is kept visible at all times — the result is a reviewable assistant.
Everything in one place
Photo & text meal analysis
Snap a photo or type a description — AI returns structured macros, glycemic index/load, and fiber with confidence labels.
Two analysis depths
Fast scan for daily logging (~2,400 tokens); deep scan with parallel blind audit and reasoning mode (~12,288 tokens).
Customizable dashboard
Composable nutrition, training, and glucose widgets with presets and a drag-and-drop editor.
Unified 24-hour timeline
Meals, glucose, activity, measurements, and plans on one continuous chart.
Connected-device integration
Dexcom G7 glucose, Google Fit, Fitbit, Android Health Connect, Nightscout, and Huawei via Health Sync.
Ma cuisine food library
Save reusable foods and dishes; quick composer and cooked-dish composer for building your own meals.
Multilingual EN / FR / ES / AR
Full RTL Arabic support, region-aware food databases, and localized nutrition guidance.
Local-first privacy
Health data stays on device. The optional account syncs only the AI allowance — never your health history.