DataFlow - Enterprise Analytics Platform

DataFlow is a production-ready autonomous data analysis system with a beautiful glassmorphism UI. Upload CSV files and get AI-powered insights, interactive charts, predictions, and comprehensive analytics.

Features

Core Capabilities

• CSV Upload and Processing - Drag and drop CSV files for instant analysis
• Auto Data Profiling - Automatic data type detection, statistics, and quality assessment
• Data Cleaning - Automatic handling of missing values, outliers, and duplicates
• Feature Engineering - Intelligent feature creation and transformation

AI-Powered Analytics

• Smart Chart Selection - AI recommends the best visualization for your data
• Automated Insights - Generate business insights using LLM (Hyperbolic API)
• Anti-Hallucination - Validates AI-generated insights against actual data
• Custom Analytics - Tableau and Power BI style free-form workspace

Visualizations

• 30+ Chart Types - Bar, line, area, pie, donut, treemap, scatter, heatmap, and more
• Interactive Charts - Zoom, pan, hover tooltips, real-time updates
• Prediction Charts - Trend projection, moving averages, growth rate forecasting
• Custom Analytics - Build custom dashboards with multiple charts

Dashboard and Reporting

• My Dashboard - Pin and arrange favorite visualizations
• PDF Export - Export dashboards and reports as PDF
• HTML Reports - Beautiful dark-themed reports with all analytics
• Currency-Aware KPIs - Support for USD, IDR, EUR, GBP, JPY, SGD

Quick Start

Prerequisites

• Docker and Docker Compose
• Python 3.12+ (for local development)
• Node.js 20+ (for frontend development)

Option 1: Docker (Recommended)

Production Mode

# Clone the repository
git clone https://github.com/Adefebrian/DataFlow.git
cd DataFlow

# Start production (nginx static build)
make up

# Or without make:
docker compose --profile prod up -d --build

Development Mode (Hot Reload)

# Start with Vite hot-reload
make dev

# Or without make:
docker compose --profile dev up -d

Access: http://localhost:3000
Login: admin / admin123

Option 2: Local Development

Backend

# Create virtual environment
python -m venv venv
source venv/bin/activate  # Linux/Mac
# or: venv\Scripts\activate  # Windows

# Install dependencies
pip install -r requirements.txt

# Copy environment file
cp .env.example .env
# Edit .env with your settings

# Run the server
python main.py

Frontend

cd frontend
npm install
npm run dev

Configuration

Environment Variables

Create a .env file based on .env.example:

Variable	Description	Default
DATABASE_URL	PostgreSQL connection string	SQLite (local)
REDIS_URL	Redis connection string	None
HYPERBOLIC_API_KEY	API key for AI insights	Required
USE_SQLITE	Use SQLite instead of PostgreSQL	true
R2_*	Cloudflare R2 storage (optional)	None

Getting a Hyperbolic API Key

1. Visit hyperbolic.xyz
2. Sign up for an account
3. Generate an API key
4. Add it to your .env file

Project Structure

DataFlow/
├── Makefile                     # Docker shortcuts
├── docker-compose.yml           # Multi-container setup
├── Dockerfile                   # Backend (Python/FastAPI)
├── requirements.txt             # Python dependencies
├── main.py                      # Entry point
├── .env.example                 # Environment template
├── .gitignore                   # Git ignore rules
│
├── src/
│   ├── api/
│   │   ├── routes.py            # Main API routes
│   │   ├── charts.py            # Chart generation
│   │   ├── insights.py          # AI insights
│   │   ├── prediction_charts.py # Forecasting
│   │   └── ...
│   ├── agents/
│   │   ├── data_profiler.py     # Data analysis
│   │   ├── feature_engineer.py  # Feature creation
│   │   ├── report_generator.py  # Report generation
│   │   └── ...
│   ├── services/
│   │   ├── llm.py               # LLM integration
│   │   ├── storage.py           # File storage
│   │   └── ...
│   └── ...
│
└── frontend/
    ├── Dockerfile               # Frontend build
    ├── nginx.conf               # Production server
    ├── vite.config.ts           # Vite configuration
    ├── package.json             # Node dependencies
    └── src/
        ├── pages/               # React pages
        ├── components/          # UI components
        ├── hooks/              # Custom hooks
        └── ...

API Endpoints

Method	Endpoint	Description
POST	/auth/login	User authentication
POST	/upload	Upload CSV file
POST	/pipeline/run	Start analysis pipeline
GET	/pipeline/{job_id}/status	Get job status
GET	/pipeline/{job_id}/analytics	Get analytics results
GET	/pipeline/all	List all jobs
GET	/health	Health check

Docker Commands

make up          # Production build
make dev         # Development mode
make down        # Stop all containers
make rebuild     # Force rebuild
make logs        # View all logs
make logs-api    # View API logs
make shell       # Shell into container

Deployment

Render.com (Recommended for Free Tier)

1. Push to GitHub
2. Connect GitHub to Render
3. Create a new Web Service
4. Settings:
   • Build Command: pip install -r requirements.txt
   • Start Command: uvicorn src.api.routes:app --host 0.0.0.0 --port $PORT

Railway

1. Install Railway CLI
2. Run: railway init
3. Run: railway up
4. Set environment variables in Railway dashboard

Fly.io

1. Install Fly CLI
2. Run: fly launch
3. Run: fly deploy

Heroku

1. Create a Procfile with:

   web: uvicorn src.api.routes:app --host 0.0.0.0 --port $PORT
   

2. Run: heroku create
3. Run: git push heroku main

Tech Stack

Backend

• FastAPI - Modern Python web framework
• Pandas - Data manipulation
• LangGraph - AI agent orchestration
• Plotly - Interactive charts
• SQLite/PostgreSQL - Database

Frontend

• React 18 - UI framework
• Vite - Build tool
• Tailwind CSS - Styling
• Plotly.js - Chart rendering
• Recharts - Additional charts

Screenshots

The platform features a modern glassmorphism design with:

• Dark theme by default
• Interactive charts with tooltips
• Drag and drop file upload
• Custom analytics workspace
• Dashboard with pinned charts

License

Personal Use Only - This software is for personal use only. Commercial use is not permitted.

Contact

For questions or inquiries:

• Instagram: @prerich.brian
• Email: adefebrianpro@gmail.com

---

Built with love using FastAPI, React, and Plotly