PROJECT 06 / Interactive prototype
Nourish
A voice-first nutrition companion, built for small daily wins.
- Problem
- Food logging becomes difficult when every entry requires exact brands and quantities.
- My contribution
- Directed voice-first logging, staged onboarding, searchable preferences, portion review, and a motivating mobile interface.
- Working today
- A front-end prototype with editable food estimates, diary, meal planning, hydration, progress, and local saves.
- Evidence
- Onboarding, meal entry, and quest progress were checked in the browser. Nutrition estimates and coaching use demo providers.
A mobile-first nutrition prototype that turns natural meal descriptions into editable estimates, with personalized setup, meal planning, and encouragement beyond the scale.
Challenge & context
Make a daily check-in feel easier to start.
Logging every brand and exact gram can turn a simple meal into an administrative task. Nourish explores a more approachable experience: describe what you ate, review a sensible estimate, and move on. The challenge was to keep that simplicity while making uncertainty and user control visible.
- Center the home screen on one clear action: “Tell me what you ate.”
- Use calories, protein, carbs, fat, and fiber as estimates rather than promises of precision.
- Support goals without framing food choices or missed days as failure.
My role & approach
Design around people, preferences, and momentum.
I directed the product concept, onboarding structure, mobile interaction design, and playful visual identity through AI-assisted development. I refined the experience into distinct setup steps and searchable selections, then carried the same green-and-white design language through the whole app.
- Separate personal details, activity and pace, food preferences, needs and goals, and final review.
- Let people search, select, and remove preference tags without losing earlier choices.
- Use tactile controls, short transitions, and reduced-motion support to keep the experience welcoming.
Solution & key workflows
Describe a meal. Review it. Build on the next small win.
The working prototype connects natural-language food entry with editable meal breakdowns, a dated nutrition ledger, and daily balance views. Speech recognition is available in supported browsers, with typed entry as a fallback. Recipe discovery, planning, grocery lists, and progress views support the routine around food logging.
- Review food choices, portions, and uncertainty ranges before an entry reaches the diary.
- Track water, movement, weight trends, weekly check-ins, and non-scale victories.
- Earn XP for check-ins, planning, and victories—not for eating less or losing weight.
Outcome & next steps
A functioning experience with a clear path to real integrations.
Nourish demonstrates the end-to-end front-end experience with fictional sample data and browser-local persistence. Onboarding selections, diary entry, and earned quest progress were checked in the browser. No weight-loss outcomes, retention improvements, or clinical effectiveness have been measured.
- Current boundaries: food interpretation uses demo averages; the coach is scripted; photos are previews; reminders store preferences.
- Next: connect speech, an LLM, and a verified nutrition database through the existing provider interfaces, with clear review and error handling.
- Before production: validate nutrition guidance, allergies, privacy, accounts, synchronization, and accessibility with appropriate experts and user testing.
