A production-ready, async file conversion platform.
Upload. Convert. Download. At scale.

---

What is SmokeByte?

SmokeByte is a backend platform that lets users securely upload files, convert them to different formats through an asynchronous background queue, track conversion status in real-time, and download the results from cloud storage — all via a clean REST API.

It's built to handle image, document, audio, and video conversions reliably, even under load.

---

Features

🔐 Authentication & Security

Feature	Details
JWT Auth	Access + Refresh token architecture
Password Hashing	bcrypt
Route Protection	Middleware-guarded endpoints
User Isolation	Each user sees only their own resources
Hardening	Helmet, CORS, Rate Limiting

🔄 File Conversion

🖼️ Images — via Sharp

From	To
JPG	PNG, WEBP
PNG	JPG, WEBP
WEBP	JPG, PNG

📄 Documents — via LibreOffice

From	To
PDF	DOCX, TXT
DOCX	PDF, TXT
TXT	PDF, DOCX
PPT / PPTX	PDF

🎵 Audio & Video — via FFmpeg

Supports cross-format conversion across: MP4, AVI, MOV, WEBM, MP3, WAV, AAC, FLAC, OGG, M4A, WMA

⚙️ Queue-Based Processing

• BullMQ powered job queue backed by Redis
• Background worker processes jobs asynchronously
• Each conversion tracked by a BullMQ Job ID
• Workers are decoupled from the API — scale independently

☁️ Storage

• Converted files stored in Supabase Object Storage
• Temporary local files auto-cleaned post-upload
• Secure, user-scoped download URLs

📦 User Features

• Conversion history per user
• Real-time job status polling
• Secure file download
• Full resource isolation between users

---

Architecture

Frontend / API Client
        │
        ▼
  Express REST API
  ┌─────┴──────────────────────┐
  │  Auth  │  Upload  │  Jobs  │
  └─────┬──────────────────────┘
        │
        ▼
    BullMQ Queue  ←──  Redis
        │
        ▼
     Worker Process
     ┌────┬────┬────┐
     │    │    │    │
     ▼    ▼    ▼    ▼
  Sharp FFmpeg LibreOffice
     │    │    │
     └────┴────┘
           │
           ▼
   Supabase Object Storage
           │
           ▼
       PostgreSQL

---

Workflow

flowchart TD
    A[User Uploads File] --> B[Upload Validation]
    B --> C[Store Original in Supabase]
    C --> D[Create BullMQ Job]
    D --> E[Worker Picks Up Job]
    E --> F{File Type?}

    F -->|Image| G[Sharp]
    F -->|Document| H[LibreOffice]
    F -->|Audio/Video| I[FFmpeg]

    G --> J[Upload Converted File]
    H --> J
    I --> J

    J --> K[Save File Record]
    K --> L[Save Conversion Log]
    L --> M[Job ID Returned to Client]
    M --> N[Client Polls /status/:jobId]
    N --> O[Client Downloads Converted File]

Loading

---

Project Structure

src/
├── config/
│   ├── db.js                 # PostgreSQL connection
│   └── supabase.js           # Supabase client
│
├── controllers/
│   ├── authController.js
│   ├── conversionController.js
│   ├── statusController.js
│   ├── historyController.js
│   └── userController.js
│
├── jobs/
│   ├── queue.js              # BullMQ queue definition
│   └── worker.js             # Background job processor
│
├── middlewares/
│   ├── authMiddleware.js     # JWT verification
│   ├── uploadMiddleware.js   # Multer file handling
│   ├── rateLimiter.js
│   └── errorHandler.js
│
├── models/
│   ├── userModel.js
│   ├── fileModel.js
│   ├── conversionLogs.js
│   └── refreshTokenModel.js
│
├── routes/
│   ├── authRoutes.js
│   ├── conversionRoutes.js
│   ├── statusRoutes.js
│   ├── historyRoutes.js
│   └── userRoutes.js
│
├── services/
│   ├── imageService.js       # Sharp logic
│   ├── documentService.js    # LibreOffice logic
│   ├── audiovideoService.js  # FFmpeg logic
│   └── storageService.js     # Supabase upload/download
│
├── utils/
│   ├── jwtUtils.js
│   ├── fileUtils.js
│   └── logger.js
│
├── app.js
└── server.js

---

API Reference

Auth

POST   /auth/register          # Create new account
POST   /auth/login             # Login, receive access + refresh tokens
POST   /auth/refresh-token     # Get new access token using refresh token
POST   /auth/logout            # Invalidate session

User

GET    /user/me                # Get authenticated user profile
GET    /user/history           # Get user's conversion history

Conversion

POST   /convert/image          # Convert image files
POST   /convert/document       # Convert document files
POST   /convert/media          # Convert audio/video files

All /convert/* endpoints return a jobId for polling.

Status & Download

GET    /status/:jobId          # Poll conversion job status
GET    /download               # Download converted file

---

Getting Started

1. Clone & Install

git clone <repository-url>
cd SmokeByte
npm install

2. Configure Environment

Create a .env file at the project root:

PORT=3000

DATABASE_URL=

JWT_SECRET=
JWT_REFRESH_SECRET=

REDIS_URL=

SUPABASE_URL=
SUPABASE_KEY=
SUPABASE_SERVICE_ROLE_KEY=
SUPABASE_BUCKET=convert-files

NODE_ENV=production

3. Run Development Server

# Start API server
npm run dev

# Start worker (separate terminal)
node src/jobs/worker.js

The API server and the worker are separate processes. Both need to be running for end-to-end conversions to work.

---

Docker

# Build image
docker build -t smokebyte .

# Run container
docker run -p 3000:3000 smokebyte

---

Tech Stack

Layer	Technology
Runtime	Node.js
Framework	Express.js
Database	PostgreSQL + Sequelize ORM
Auth	JWT, Refresh Tokens, bcrypt
Queue	BullMQ + Redis (Upstash)
Storage	Supabase Object Storage
Image Processing	Sharp
Media Processing	FFmpeg
Document Processing	LibreOffice
Infrastructure	Docker, PM2, Cloudflare Tunnel

---

Roadmap

• Email verification on signup
• Password reset via email
• Refresh token rotation
• Admin dashboard
• Conversion analytics
• WebSocket-based real-time status updates
• Multi-file batch conversion

---

Author

Pushpak Pathe

Backend Developer — Node.js · Express.js · PostgreSQL · Redis · Docker

---

Built with ☕ and way too many conversion edge cases.