☕ DevBrew

Coffee chats for developers. Get matched 1:1 with another dev who shares your interests. 15 minutes of real conversation. Then back to work.

---

✨ Features

• 🎯 Interest-Based Matching — Pick your tech interests (React, Kubernetes, AI, etc.) and get matched with developers who share them
• 📹 HD Video Chat — Real-time peer-to-peer video calls powered by WebRTC
• ⚡ Instant Connect — No signup required. Jump straight into conversations
• 🔀 Skip & Match — Not vibing? Skip and instantly find a new partner
• 🔒 Privacy First — No data stored. Connections are ephemeral

---

🏗️ Tech Stack

Layer	Technology
Frontend	Next.js 16, React 19, TypeScript
Real-time	Socket.IO for signaling
Video	WebRTC with STUN/TURN servers
Backend	Node.js, Express, TypeScript
Containerization	Docker, Docker Compose, Nginx

---

🚀 Getting Started

Prerequisites

• Node.js 18+
• npm or yarn
• Docker & Docker Compose (for containerized setup)

---

🐳 Docker

The entire stack is containerized with production-grade Docker practices.

Architecture

flowchart LR
    U[Client Browser] -->|HTTP :80| N[Nginx Reverse Proxy]

    subgraph DN[Docker Network]
        N -->|/| F[Frontend<br/>Next.js :3000]
        N -->|/socket.io| B[Backend<br/>Express :3002]
    end

Loading

Quick Start

# Clone and run — that's it
git clone https://github.com/yashpatil/DevBrew.git
cd DevBrew
make dev

Open http://localhost and you're live.

Make Targets

Command	Description
make dev	Build & start all services (attached)
make prod	Build & start all services (detached)
make build	Build all Docker images
make down	Stop all containers
make clean	Stop, remove images, volumes & orphans
make logs	Tail logs from all services
make ps	Show running containers

Docker Highlights

• Multi-stage builds — 3-stage Dockerfiles for both frontend and backend keep final images minimal
• Non-root users — All containers run as unprivileged users
• Health checks — Every service has a health check; depends_on uses condition: service_healthy
• Layer caching — package.json copied before source for optimal cache hits
• Resource limits — CPU and memory constraints on every service
• Nginx reverse proxy — Routes traffic, upgrades WebSocket connections, gzip compression, security headers
• Standalone Next.js — output: "standalone" produces a self-contained server (~50MB vs ~500MB)

---

📁 Project Structure

devbrew/
├── backend/
│   ├── src/
│   │   └── index.ts          # WebRTC signaling server
│   ├── Dockerfile            # Multi-stage build
│   ├── .dockerignore
│   ├── package.json
│   └── tsconfig.json
│
├── frontend/
│   ├── app/
│   │   ├── page.tsx          # Landing page
│   │   ├── interests/
│   │   │   └── page.tsx      # Interest selection
│   │   ├── chat/
│   │   │   └── page.tsx      # Video chat room
│   │   ├── globals.css       # Global styles
│   │   └── layout.tsx        # Root layout
│   ├── Dockerfile            # Multi-stage build
│   ├── .dockerignore
│   ├── package.json
│   └── tsconfig.json
│
├── nginx/
│   ├── nginx.conf            # Reverse proxy config
│   └── Dockerfile
│
├── docker-compose.yml        # Full stack orchestration
├── Makefile                  # Developer shortcuts
└── README.md

---

⚙️ How It Works

1. User joins → Selects interests → Joins the matching queue
2. Matching algorithm → Finds best match based on shared interests
3. WebRTC handshake → Signaling server facilitates offer/answer exchange
4. Peer connection → Direct video/audio stream between users
5. Chat ends → User can skip or leave; no data is stored

sequenceDiagram
    participant A as User A (Peer A)
    participant S as Signaling Server
    participant B as User B (Peer B)

    A ->> S: Socket.IO
    S ->> A: Socket.IO
    S ->> B: Socket.IO
    B ->> S: Socket.IO

    Note over A,B: WebRTC Peer Connection (Direct Video/Audio)

Loading

---

🔧 Configuration

STUN/TURN Servers

The app uses public STUN servers and OpenRelay TURN servers for NAT traversal:

const ICE_SERVERS = {
    iceServers: [
        { urls: 'stun:stun.l.google.com:19302' },
        { urls: 'stun:stun1.l.google.com:19302' },
        {
            urls: 'turn:openrelay.metered.ca:80',
            username: 'openrelayproject',
            credential: 'openrelayproject',
        },
    ],
};

For production, consider using your own TURN server (e.g., Coturn) for reliability.

---

📦 Deployment

With Docker (Recommended)

make prod

Update NEXT_PUBLIC_SIGNALING_SERVER build arg in docker-compose.yml to your production backend URL.

Without Docker

Backend

cd backend
npm run build
npm start

Frontend

cd frontend
npm run build
npm start

---

🤝 Contributing

Contributions are welcome! Feel free to:

• Report bugs
• Suggest features
• Submit pull requests

---

☕ Grab a coffee. Find your people.