Vision
Build the most frictionless calorie tracking app for serious nutrition users.
The product is a memory-first nutrition decision system. Exact workflows remain first class when exact data exists. Assistance creates reviewable drafts when uncertainty exists. Recommendations become more useful as the user builds trustworthy history.
Product principles
- Frictionless logging
- Memory over prediction
- Context beats search
- Capture once, reuse many times
- Show progress, not just numbers
- Explainable recommendations
- Progressive disclosure
- Privacy by default
- Celebrate consistency over perfection
- Holistic health
- User owns their data
- Reduce uncertainty
- Protect data integrity over data volume
Core jobs to be done
Daily
- Track food with minimal friction.
- Understand today’s calorie and protein status.
- Know the next best action.
Weekly
- Understand whether the plan is working.
- Validate incomplete data before algorithm adjustments.
Long-term
- Discover behavioural patterns.
- Connect body changes with nutrition without presenting correlation as proof.
Onboarding flow
01Sign in and confirm body data
02Select lose, maintain, gain, or recomposition
03Connect health data or set activity
04Confirm target and log the first meal
Today command centre
The home screen puts Today first: show where the person is now, keep the diary visible, and make start or continue tracking obvious.
- calories remaining;
- protein remaining;
- daily progress;
- every logged item and per-meal total;
- start or continue tracking;
- one suggested next action;
- one helpful insight that can be disabled.
A compact “This week so far” layer is secondary. It shows clearly labelled weekly averages or accumulated totals, data coverage, and progress toward the person’s weekly intent. Daily body signals, nutrition, alcohol, steps, training, and notes can roll up into this view. Missing days are never zero, and no individual day receives a moralized score.
Daily logging
Entry points are repeat usual meal, voice, barcode scanner, and search. Every confirmed entry creates an attributable Food Log Event.
People can group events without rewriting history and optionally save reusable Meal Templates. In a shared prepared meal, each person logs their own portion without changing the other person’s history or targets.
Weekly reflection
- Celebrate consistency without judging individual food choices.
- Show a compact weekly snapshot.
- Identify suspicious or incomplete days.
- Let the user correct or classify each day.
- Add photos and measurements when useful.
- Explain what changed, which evidence was used, and how confident the system is.
- Offer one recommendation to accept, defer, or reject.
Pattern detection and future
Future patterns may identify that Fridays are often incomplete or that a high-protein breakfast correlates with more successful days. The product must distinguish observations from possible explanations.
Later opportunities include coach collaboration, MCP integrations, and deeper AI coaching. These should follow proof that the daily and weekly loops retain serious users.
Source and context
The Markdown file is the living source. The competitor report tests the product thesis against real user evidence; the commercial analysis tests whether the position can support a business; the flows translate the product into complete sequences.