FalaFit
A nutrition tracking app that replaces long forms with a conversation. Users speak or type what they ate, and the system structures foods, calculates macronutrients and updates daily progress.
Reduce friction without losing structure
Food tracking often requires searches, measurements, portions and multiple fields. That effort appears precisely when consistency matters, making the habit difficult to maintain.
FalaFit had to accept natural language without turning the experience into a black box. Input needed to be simple while the result still produced editable foods, amounts and macronutrients that were useful for tracking.
Voice in, organized data out
The user records up to 60 seconds of audio or sends text. An Edge Function validates the session and controls usage before forwarding the content to the AI workflow.
Audio is transcribed, foods are extracted and normalized, and a local knowledge base is checked before any new estimate. The result returns as a structured meal, updates daily progress and remains editable.
The experience ships as an installable web application and as native iOS and Android apps, sharing the same data domain and access controls.
How it works
Natural input
Audio or text describes a meal without a long form.
Validation
The edge function validates session, origin and usage limits.
Interpretation
Whisper transcribes and Claude turns speech into food items.
Normalization
The food cache reduces repetition and keeps data reusable.
Tracking
Macros, daily progress and history update in the interface.
Product and engineering decisions
Offline first
A persisted queue accepts entries on unstable connections and synchronizes when the network returns.
Cache before AI
Known foods are reused. AI handles missing references, reducing cost and inconsistency.
Layered security
PKCE authentication, table-level RLS, authenticated webhooks and usage limits protect data and infrastructure.
Application interfaces
What reached production
- Published apps on the App Store and Google Play
- Voice and text logging with calorie and macronutrient calculations
- Offline queue, favorite meals and quick macro editing
- Weight, measurements, photos and nutrition progress tracking
- Nutritionist and patient workflow with controlled invitations
- Automated mobile build and distribution pipeline
Luis Braghin
Luis Braghin · Product, software and automation development