🔥 Social Profile Roaster (Linkedin & Github)

A brutally honest, dual-platform roasting tool. This app uses AI to generate witty, savage roasts based on your professional presence - whether you're a "Corporate LinkedIn Guru" or a "GitHub Green Square Addict"

New Features

• Dual-Mode UI - Brand-matching themes for both LinkedIn (Corporate Blue) and GitHub (Dev Green) with a unified Design System
• AI-Powered Roasting - Uses Groq's LLaMA 3.3 model to generate brutal, witty roasts
• LinkedIn Profile Scraping - Automatically fetches and analyzes LinkedIn profiles
• Fast & Efficient - Cached authentication for quick roast generation
• Personal Roasts - Each roast is tailored to specific profile details, not generic

Tech Stack

• Frontend: Next.js 16 + React 19 + Tailwind CSS
• Backend: Next.js API Routes
• AI: Groq API (llama-3.3-70b-versatile)
• Animation: Native CSS animations (keyframes)
• Deployment: Netlify (with GitHub integration)

Navigation Structure

The app is now structured to support parallel platform experiences:

• /linkedin - The Corporate Grind Roast
• /github - The Code Monkey Roast

Getting Started

Prerequisites

• Node.js 18+
• Groq API key (get one at https://console.groq.com)

Installation

1. Clone the repository:

git clone https://github.com/aiman-mumtaz/profile-roast.git
cd profile-roast

2. Install dependencies:
3. Install dependencies:

npm install

3. Set up environment variables in .env.local:

GROQ_API_KEY=your_groq_api_key_here

4. Run the development server:

npm run dev

5. Open http://localhost:3000 in your browser

How It Works

1. User pastes a LinkedIn profile URL
2. App logs into LinkedIn and scrapes the profile data
3. Profile text is sent to Groq's API with a custom prompt
4. AI generates a brutal, personalized roast
5. Roast is displayed with smooth animations

API Endpoints

POST /api/roast

Generate a roast for a LinkedIn profile or custom text.

Request:

{
  "profile": "https://www.linkedin.com/in/username"
}

Response:

{
  "roast": "Your brutal roast here..."
}

Configuration

• Max tokens: 200 (keeps roasts short and punchy)
• Temperature: 0.7 (balanced creativity vs. consistency)
• Model: llama-3.3-70b-versatile
• Browser timeout: 15 seconds for scraping

Performance Notes

• API response time: ~3-5 seconds
• No authentication delays (no scraping required)
• Lightweight, privacy-respecting architecturefiles
• No hate speech, slurs, or protected class attacks

Legal & Ethical

• Only scrape public LinkedIn profiles
• No hate speech, slurs, or protected class attacks
• Roasts are satirical and comedic in nature
• Use responsibly and for entertainment purposes only

Future Improvements

• Support for other social profiles (Github Added)
• Custom roast intensity levels
• Share roasts on social media
• Roast history/favorites
• Multi-language support
• Analytics dashboard
• Export roasts as high-quality PNGs for sharing

License

MIT

Support

For issues or questions, please open an issue on the repository.

---

Built with ❤️ and brutality 🔥