Chihuahua|
Projects

Robfit.co — Fitness & Nutrition Coaching Platform

Coaching platform for a real fitness coach and his clients: routines, AI macro tracking from text or photos, habits, progress photos and chat, on the web and in an Android app.

Andrés Aguilar
Andrés Aguilar
September 16, 2026
  • nextjs
  • react
  • supabase
  • flutter
  • openrouter
  • ai
  • cloudflare-r2
Visit projectrobfit.vercel.app
Robfit.co — Fitness & Nutrition Coaching Platform - Image 1
Robfit.co — Fitness & Nutrition Coaching Platform - Image 2
Robfit.co — Fitness & Nutrition Coaching Platform - Image 3
Robfit.co — Fitness & Nutrition Coaching Platform - Image 4

Overview

Robfit.co is the coaching platform of Roberto, a real fitness and nutrition coach, and his clients. It brings training, diet and habits into one system built on three surfaces that share the same API and database: a public landing page, a client dashboard and an admin panel for the coach, plus an Android app built with Flutter.

Features

  • Guided onboarding — clients sign up (email or Google), fill in their profile, health data, activity level and goal, and upload guided progress photos
  • Routines — the coach assigns routines with sets, reps and rest times; clients mark completed workouts
  • AI nutrition agent — describe a meal or send a photo and the agent estimates calories, protein, carbs and fat, logs it and answers questions about macros, habits and progress
  • Macro tracking — daily targets set by the coach, a macro calendar and logging of forgotten meals from up to 7 days back
  • Habits and progress — habit check-ins, weight, BMI, body measurements and lifts over time, plus a shareable recap
  • Chat with the coach — a human channel with Roberto next to the AI agent channel
  • Coach panel — per-client editors for profile, routine, macros and habits, data capture, and moderation of meals and photos
  • Android app — distributed outside the Play Store from a download page with a QR code and an in-app update prompt

Technical Details

  • Web: Next.js 16 and React 19 with Server Components, Server Actions and route handlers shared with the mobile app; plain CSS with a brand design system
  • Backend: Supabase (Postgres + Auth) with Row Level Security as the real security layer, including triggers and policies that lock identity and role columns
  • AI: OpenRouter (OpenAI-compatible API) with JSON Schema structured output, model fallbacks, timeouts, retries and a per-user daily limit; idempotent meal logging with client tokens
  • Storage: private Cloudflare R2 bucket with short-lived signed URLs for photos and the APK
  • Mobile: Flutter Android app consuming the same API with bearer tokens
  • Deployment: Vercel with a daily cron for diet reminders