🕵️‍♂️ Reel Scout

Find any lost Instagram Reel. Describe it. Done.

🚀 Live Demo: Try Reel Scout Here

Instagram's native search is broken. If you forgot to save a Reel, it’s practically gone. **Reel Scout** fixes this. Type a messy, natural language description of what you saw, and my custom-built search engine will hunt it down.

---

✨ Features

• 🧠 Proprietary Query Architect: A custom Python engine that parses messy English into highly targeted Google Dorks.
• ⚡ SerpAPI Integration: Bypasses basic limits by routing the generated dorks through SerpAPI's robust Google Search infrastructure.
• 🎨 Stunning Glassmorphism UI: Animated CSS orbs, glowing search inputs, scan-line loading animations, and backdrop blur cards. Dark mode, purple/pink IG aesthetic.
• شف Transparent Engine: Shows the exact Google Dork generated in the UI, keeping the background process transparent and educational.
• 📱 Fully Responsive: Beautiful on mobile, tablet, and desktop.
• 🛠️ Zero Frontend Build: A single index.html file using Tailwind via CDN. No npm, no bloat.

⚙️ How It Works

🧑 [User Input] 
 "cat dancing on rooftop in Mumbai #viral #dance"
      │
      ▼
🏗️ [Query Architect] (Python Engine / Gemini)
 Strips noise, identifies entities, formats exact syntax
      │
      ▼
🔍 [Google Dork Generated] 
 site:instagram.com/reel/ OR site:instagram.com/reels/ "cat dancing" "Mumbai" #viral #dance
      │
      ▼
🌐 [SerpAPI Google Search]
 Executes the precise Dork against indexed Instagram pages
      │
      ▼
✨ [Reel Scout UI]
 Parses JSON, drops the noise, displays glowing Glassmorphism cards

🏗️ Project Structure

Reel Scout is designed to be lean and modular:

• Backend: Python 3.12 + Flask 3.0. Handles the API routing, the custom Query Architect logic, and secure HTTP calls to SerpAPI.
• Frontend: Pure HTML5 + Vanilla JS + Tailwind CSS. Designed as a Single Page Application (SPA) without the overhead of heavy frameworks.
• Search Engine: SerpAPI + Gemini AI.

📂 Folder Structure

reel-scout/
├── app.py               # Flask backend, routes, query builder, search
├── requirements.txt     # flask, requests, python-dotenv, google-genai
├── vercel.json          # Deployment configuration
├── .env.example         # API key template
├── .gitignore           # Keeps .env out of Git
└── templates/
    └── index.html       # Complete frontend (Tailwind + Vanilla JS)

🚀 Local Evaluation Guide

You can run Reel Scout locally to evaluate the codebase and architecture.

1. Clone the repository: git clone https://github.com/yourusername/reel-scout.git cd reel-scout

2. Create and activate a virtual environment: python -m venv venv venv\Scripts\activate

3. Install dependencies: pip install -r requirements.txt

4. Configure your environment variables:

   • Rename the .env.example file to .env
   • Add your API keys (instructions below).

5. Run the local server: python app.py (Open http://localhost:5000 in your browser.)

🔑 Getting API Keys

To test the application locally, you'll need two API keys. The app will throw a RuntimeError if these are missing.

1. Get the Gemini API Key:

   • Go to Google AI Studio (aistudio.google.com).
   • Sign in and click "Get API key" -> "Create API key".
   • Copy the generated key.

2. Get the SerpAPI Key:

   • Go to serpapi.com and create a free account.
   • Navigate to your Dashboard.
   • Copy "Your Private API Key".

3. Update your .env file: GEMINI_API_KEY=your_gemini_key_here SERP_API_KEY=your_serpapi_key_here

🌍 Deployment Policy

While the source code is provided under the MIT License for educational purposes and portfolio showcasing, please do not deploy your own public hosted instance of this application using my exact UI and branding. Reel Scout is actively deployed as a personal project, and creating mirror deployments causes API rate-limiting and duplication issues.

🤝 Contributions & Forks

As this is primarily a portfolio piece, pull requests for new features are currently not being accepted. You are welcome to star the repository or fork it strictly for private reading and code review.

⚖️ Copyright & License

© 2026 Aditya Lande. Source code is licensed under the MIT License.

🧑‍💻 The Architect

Aditya Lande — First-year Computer Engineering student at Don Bosco Institute of Technology (DBIT), Mumbai.