Picks - AI-Powered Movie & TV Recommendations

A minimalist, AI-powered movie and TV show recommendation app that helps you discover your next favorite watch based on your mood, preferences, and viewing history.

✨ Features

• 🎭 Mood-Based Quick Picks: Get instant recommendations based on how you're feeling (10 different moods)
• 🤖 AI-Powered Search: Advanced RAG (Retrieval-Augmented Generation) workflow using Google Gemini
• 📝 Personal Watchlist: Save recommendations with local storage and export functionality
• 🔍 Smart Filters: Quick genre and length filters for better recommendations
• 🌐 Web Search Integration: Fallback to real-time web search for latest content
• 📱 Responsive Design: Beautiful, minimal UI that works on all devices

🚀 Getting Started

Prerequisites

• Node.js 18+ or Bun
• Google Generative AI API key
• Supabase account (for vector database)
• Tavily API key (for web search)

Installation

1. Clone the repository:

git clone <your-repo-url>
cd picks

2. Install dependencies:

bun install
# or
npm install

3. Set up environment variables:

cp .env.example .env.local

Add your API keys to .env.local:

GOOGLE_GENERATIVE_AI_API_KEY=your_gemini_api_key
NEXT_PUBLIC_SUPABASE_URL=your_supabase_url
NEXT_PUBLIC_SUPABASE_ANON_KEY=your_supabase_anon_key
TAVILY_API_KEY=your_tavily_api_key

4. Run the development server:

bun dev
# or
npm run dev

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

🏗️ Tech Stack

• Framework: Next.js 14 (App Router)
• Styling: Tailwind CSS + Radix UI
• AI/ML: Google Gemini 2.0 Flash, LangChain, LangGraph
• Database: Supabase (Vector Store)
• Search: Tavily API
• State Management: React Hooks + Local Storage
• Type Safety: TypeScript

📖 How It Works

AI-Powered Recommendations

The app uses an advanced agentic RAG workflow:

1. Query Processing: User input is processed and optimized
2. Vector Search: Semantic search through movie/TV database
3. Relevance Grading: AI evaluates if results match the query
4. Web Search Fallback: If local results aren't relevant, searches the web
5. Response Generation: AI generates personalized recommendations

Mood-Based Quick Picks

Choose from 10 distinct moods:

• 😊 Happy & Uplifting
• 🤩 Thrilled & Energetic
• 😌 Calm & Relaxed
• 🥰 Romantic & Loving
• 🤔 Curious & Intrigued
• 🥺 Nostalgic & Sentimental
• 🧗 Adventurous & Bold
• 🧠 Thoughtful & Cerebral
• 🏠 Cozy & Comfortable
• 🎭 Emotional & Dramatic

Watchlist Features

• Local Storage: All data stored locally for privacy
• Export Options: Download your watchlist as JSON
• Status Tracking: Mark items as "Want to Watch", "Watching", or "Watched"
• Search & Filter: Find items in your watchlist quickly
• Cross-tab Sync: Changes sync across browser tabs

🎯 Free Tier Optimized

The app is optimized for Google Gemini's free tier:

• Token Management: Automatic token usage tracking
• Rate Limiting: Built-in protection against API limits
• Efficient Prompts: Minimized token consumption
• Smart Caching: Reduced redundant API calls

🤝 Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

📄 License

This project is open source and available under the MIT License.

🙏 Acknowledgments

• Built with Next.js
• AI powered by Google Gemini
• UI components from Radix UI
• Styled with Tailwind CSS