Reflecta - AI-Powered Personal Reflection Journal

Reflecta is an intelligent journaling application that helps users with self-reflection, goal setting, and personal growth through AI-powered conversations and insights.

Live Demo

Frontend (Vercel): https://aidiary-reflecta.vercel.app/demo

Backend API (Google Cloud Run): https://reflecta-backend-762303020827.us-central1.run.app

Overview

Reflecta combines the power of AI with structured journaling to help users:

• Engage in meaningful self-reflection through AI-powered conversations
• Set and track goals using the Mandalart goal-setting framework
• Convert conversations into structured journal entries
• Track progress toward personal growth objectives
• Visualize their reflection journey through interactive timelines

Features

1. AI Reflection Assistant

• Real-time chat interface powered by OpenAI GPT
• Contextual responses based on user's goals and past reflections
• Smart conversation-to-diary conversion
• Automatic goal mapping and progress tracking

2. Goal Setting (Mandalart Framework)

• Interactive 9x9 Mandalart grid for structured goal planning
• Core goal and 8 sub-goal categories
• Each sub-goal broken down into 8 actionable tasks
• Visual progress tracking for all goals

3. Journal Management

• Create, edit, and organize journal entries
• Tag entries with moods and categories
• AI-generated entries from chat conversations
• Search and filter capabilities

4. Progress Tracking

• Visual timeline of journal entries
• Goal completion metrics
• Mood patterns and insights
• Weekly, monthly, and yearly summaries

Technology Stack

Frontend

• Framework: React.js
• Styling: CSS3 with responsive design
• State Management: React Context API
• Routing: React Router v6
• Deployment: Vercel

Backend

• Runtime: Node.js with Express.js
• Database: MongoDB Atlas
• Authentication: JWT (JSON Web Tokens)
• AI Integration: OpenAI GPT API
• Deployment: Google Cloud Run (MLOps course application)

Architecture

┌─────────────────────────────────────────────────────────────┐
│                    Frontend (Vercel)                         │
│                  React.js Application                        │
│  https://aidiary-reflecta.vercel.app                        │
└────────────────────────┬────────────────────────────────────┘
                         │
                         │ HTTPS/REST API
                         │
┌────────────────────────▼────────────────────────────────────┐
│              Backend (Google Cloud Run)                      │
│                  Express.js Server                           │
│  https://reflecta-backend-762303020827.us-central1.run.app  │
└────────────┬───────────────────────┬────────────────────────┘
             │                       │
             │ MongoDB Driver        │ OpenAI API
             │                       │
┌────────────▼────────────┐  ┌──────▼─────────────────────────┐
│   MongoDB Atlas          │  │   OpenAI GPT-4                 │
│   User Data & Journals   │  │   AI Conversation Engine       │
└──────────────────────────┘  └────────────────────────────────┘

Project Structure

reflecta/
├── reflecta-frontend/          # React frontend application
│   ├── src/
│   │   ├── components/         # React components
│   │   ├── contexts/           # Context providers (Auth, Journal)
│   │   ├── services/           # API service layer
│   │   └── App.js              # Main application component
│   ├── public/                 # Static assets
│   └── package.json
│
├── reflecta-backend/           # Node.js backend API
│   ├── models/                 # MongoDB schemas
│   │   ├── User.js
│   │   ├── Goal.js
│   │   ├── JournalEntry.js
│   │   ├── ChatSession.js
│   │   └── GoalProgress.js
│   ├── server.js               # Express server & API routes
│   ├── Dockerfile              # Container configuration
│   └── package.json
│
└── README.md                   # This file

Deployment

Frontend Deployment (Vercel)

The frontend is deployed on Vercel with the following configuration:

• Root Directory: reflecta-frontend
• Framework: Create React App
• Build Command: npm run build
• Output Directory: build

Environment variables configured in Vercel:

REACT_APP_API_URL=https://reflecta-backend-762303020827.us-central1.run.app/api
REACT_APP_APP_NAME=Reflecta
REACT_APP_APP_VERSION=1.0.0
REACT_APP_MOCK_API=false

Backend Deployment (Google Cloud Run)

The backend is deployed on Google Cloud Run, applying concepts learned from the MLOps course:

Key MLOps Practices Applied:

• Containerization: Using Docker for consistent deployment environments
• Cloud-Native Deployment: Serverless architecture with automatic scaling
• Environment Management: Secure handling of secrets and configurations
• CI/CD Ready: Source-based deployment with Cloud Build
• Monitoring: Built-in logging and health check endpoints

Deployment Command:

gcloud run deploy reflecta-backend \
  --source . \
  --region us-central1 \
  --platform managed \
  --allow-unauthenticated \
  --env-vars-file .env.yaml \
  --port 5000

Environment Variables:

• MONGODB_URI: MongoDB Atlas connection string
• JWT_SECRET: Secret key for authentication
• CORS_ORIGIN: Allowed frontend origins
• OPENAI_API_KEY: OpenAI API key
• OPENAI_API_URL: OpenAI API endpoint

Local Development

Prerequisites

• Node.js 18 or higher
• MongoDB (local or Atlas)
• OpenAI API key

Frontend Setup

cd reflecta-frontend
npm install
cp .env.example .env
# Edit .env with your configuration
npm start

The frontend will run on http://localhost:3000

Backend Setup

cd reflecta-backend
npm install
cp env.example .env
# Edit .env with your configuration
npm start

The backend will run on http://localhost:5000

API Endpoints

Authentication

• POST /api/auth/register - Register new user
• POST /api/auth/login - Login user

Goals

• GET /api/goals - Get user's goals
• POST /api/goals - Save/update goals
• POST /api/goals/ai-suggestions - Get AI-powered goal suggestions

Journal

• GET /api/journal - Get all journal entries
• GET /api/journal/:id - Get specific entry
• POST /api/journal - Create new entry

Chat

• GET /api/chat - Get chat session
• POST /api/chat - Save chat session
• POST /api/chat/enhanced - Send message with AI response

Conversion & Analysis

• POST /api/convert-to-diary - Convert conversation to journal entry
• POST /api/analyze-goal-mapping - Analyze diary for goal connections

Goal Progress

• GET /api/goals/:id/progress - Get goal progress
• POST /api/goals/:id/progress - Update goal progress
• GET /api/goals/:id/progress/summary - Get progress summary

Health Check

• GET /api/health - Backend health status

Key Features Implementation

1. AI Conversation to Diary

Users can select messages from their chat with the AI assistant and convert them into structured journal entries. The AI analyzes the conversation and:

• Generates a coherent diary entry
• Identifies related goals
• Suggests relevant tags and mood

2. Mandalart Goal Framework

The application implements the Mandalart (9x9) goal-setting method:

• 1 central goal
• 8 sub-goals around the center
• Each sub-goal expands into 8 actionable tasks (total 64 tasks)
• Visual progress tracking for all levels

3. Smart Goal Mapping

AI automatically analyzes journal entries to:

• Identify connections to existing goals
• Track progress on specific tasks
• Provide insights on goal completion patterns

Security

• JWT-based authentication
• Secure password hashing with bcrypt
• CORS configuration for frontend-backend communication
• Environment-based secret management
• MongoDB Atlas network security

Future Enhancements

• Mobile app (React Native)
• Data export functionality (PDF, JSON)
• Social features (share reflections with trusted users)
• Advanced analytics dashboard
• Voice-to-text journal entries
• Habit tracking integration
• Reminder and notification system

Credits

Developed by: Hunjun Shin

Course: MLOps - Applied cloud deployment practices using Google Cloud Run

Technologies: React, Node.js, MongoDB, OpenAI GPT, Vercel, Google Cloud Run

License

This project is developed for educational purposes.

---

Start your reflection journey today: https://aidiary-reflecta.vercel.app