FitCoach AI

FitCoach AI is a full-stack workout tracking web app built for the 24-hour take-home assignment. It lets a user log workouts, review workout history and stats, receive AI-generated motivation grounded in real activity data, and chat with an AI fitness coach.

Assignment Summary

The assignment asked for a web app where a user can:

1. Log daily workouts
2. View workout history and basic stats
3. Get AI-generated personalized motivational messages based on actual workout data
4. Ask fitness-related questions to an AI chatbot

This repository implements those core requirements with a Next.js frontend, Express backend, MongoDB persistence, Firebase authentication, and OpenRouter-powered AI responses.

Tech Stack

• Frontend: Next.js 16, React 19, TypeScript, Tailwind CSS v4
• Backend: Node.js, Express 5, Mongoose
• Database: MongoDB Atlas
• Authentication: Firebase Auth + Firebase Admin
• AI integration: OpenRouter LLM API
• Deployment target:
  • Frontend: Vercel
  • Backend: Render
  • Database: MongoDB Atlas

Requirement Coverage

1. Workout Logging

Implemented:

• Workout form with 5 activity options:
  • Running
  • Yoga
  • Cycling
  • Gym
  • Swimming
• Duration in minutes
• Date and time input with the current time as default
• Past dates allowed
• Workout data saved in MongoDB
• Logged workouts shown in reverse chronological order
• Paginated workout history API and UI

2. Stats Dashboard

Implemented stats based on logged workouts:

• Current streak
• Workouts this week
• Most frequent activity
• Total workout minutes

Behavior:

• Dashboard data refreshes immediately after a workout is logged
• No full page refresh is required

3. AI Motivational Nudge

Implemented:

• Backend endpoint to generate a motivation message from actual user workout data
• Prompt context includes streak, workout frequency, last workout timing, activity mix, and consistency patterns
• Daily motivation is stored and reused for the current day

Important note:

• The assignment mentions a "Get AI Motivation" button
• In this implementation, the motivation is automatically generated and shown on the dashboard instead of requiring a separate click
• The message is still personalized using the user's real data, which was the main scoring requirement

4. AI Q&A Chatbot

Implemented:

• Chat UI for user fitness questions
• User message plus recent chat history sent as AI context
• Scrollable conversation in the UI
• Chat history persisted in MongoDB so it survives refresh

Bonus Features Included

• Mobile-responsive UI
• Persisted chat history
• Coach tone selection
• Real-time dashboard refresh after logging workouts
• Loading and error states
• Gamification extras: points, badges, leaderboard, tournament view

Product Features In This Repo

• Firebase email/password and Google sign-in
• First-login profile setup
• Unique username validation
• Avatar upload or preset avatar support
• Weekly goal tracking
• Coach tone preference
• Daily motivation card
• AI coach chat
• Workout history tab
• Rank and leaderboard UI

Project Structure

workout/
|-- backend/
|   |-- controllers/
|   |-- middleware/
|   |-- models/
|   |-- routes/
|   |-- utils/
|   |-- app.js
|   |-- server.js
|   `-- build.sh
|-- frontend/
|   |-- public/
|   |-- src/app/
|   |-- src/components/
|   |-- src/context/
|   |-- src/lib/
|   `-- src/types/
`-- README.md

API Overview

All protected routes require a Firebase ID token in:

Authorization: Bearer <token>

Available endpoints:

• GET /api/health
• GET /api/users/me
• POST /api/users/profile
• POST /api/workouts
• GET /api/workouts?page=1&limit=10
• GET /api/stats
• POST /api/ai/motivation
• POST /api/chat
• GET /api/chat/history

Local Setup

1. Backend

Create backend/.env and add:

PORT=5000
MONGO_URI=
FIREBASE_PROJECT_ID=
FIREBASE_PRIVATE_KEY=
FIREBASE_CLIENT_EMAIL=
OPENROUTER_API_KEY=

Install and run:

cd backend
npm install
npm run dev

2. Frontend

Create frontend/.env.local and add:

NEXT_PUBLIC_FIREBASE_CONFIG={"apiKey":"...","authDomain":"...","projectId":"...","storageBucket":"...","messagingSenderId":"...","appId":"..."}
NEXT_PUBLIC_API_BASE_URL=http://localhost:5000/api

Install and run:

cd frontend
npm install
npm run dev

Open:

http://localhost:3000

Deployment

Expected Deployment Setup

• Frontend: Vercel
• Backend: Render
• Database: MongoDB Atlas

Live Deployment Links

• Frontend app: TBD
• Backend API: TBD
• GitHub repository: TBD

Live Demo Video

• Repository demo video file: frontend/Video/working_video.mp4
• Walkthrough video link: TBD
• Optional short product demo link: TBD

Deployment and video links are still placeholders because publishing credentials and final hosted URLs were not available in this workspace.

Design Decisions And Tradeoffs

• MongoDB was used instead of Supabase because the current backend is structured around Express + Mongoose and this kept development faster within the time limit.
• Daily motivation is persisted per user per day to avoid repeated LLM calls and to keep the dashboard stable during a single day.
• Avatar uploads are stored as data URLs for simplicity in the assignment scope instead of using object storage.
• The AI motivation flow is automatic on dashboard load instead of being behind a separate button. This is a UI tradeoff, but the generated content still uses real workout context.
• The backend includes fallback behavior so the app remains usable during local development if the AI provider is unavailable.

Verification

Completed locally:

• Frontend: npm run lint
• Frontend: npm run build
• Backend: node --check on backend source files