ABZanganeh/movie-agent-service

🎬 Movie Agent - AI Movie Discovery & Quiz Assistant

An intelligent AI-powered movie agent that helps you discover movies, analyze posters, play interactive quizzes, and explore movie statistics. Built with LangChain, OpenAI, and Groq.

✨ Features

• —🎯 Movie Discovery: Search and discover movies based on your preferences
• —🖼️ Poster Analysis: Upload movie posters and get AI-powered analysis with mood detection
• —🧠 Interactive Quizzes: Play movie quizzes about cast, directors, and release years
• —📊 Movie Statistics: Get insights on ratings, genres, and top-rated movies
• —💬 Conversational AI: Natural language interaction powered by LLMs (OpenAI/Groq)

🚀 Quick Start

1. 1.Set up API Keys: Configure your API keys in the Space settings:
2. 2.GROQ_API_KEY or OPENAI_API_KEY for LLM
3. 3.OPENAI_API_KEY for embeddings (required)

1. 1.Start Chatting: Ask questions like:
2. 2."Show me the best movies from 2010"
3. 3."Let's play a cast quiz"
4. 4."What are the top rated action movies?"

1. 1.Analyze Posters: Upload a movie poster image to get AI-powered analysis

🛠️ Configuration

The app uses environment variables for configuration. Set these in your Space settings:

• —LLM_PROVIDER: groq or openai (default: groq)
• —LLM_MODEL: Model name (default: llama-3.1-8b-instant)
• —GROQ_API_KEY: Your Groq API key
• —OPENAI_API_KEY: Your OpenAI API key (for embeddings and LLM if using OpenAI)
• —ENABLE_VISION: Enable poster analysis (default: false)
• —ENABLE_MEMORY: Enable conversation memory (default: true)

📚 Use Cases

• —Movie Discovery: Find movies by genre, year, rating, or director
• —Poster Analysis: Identify movies from poster images
• —Interactive Learning: Play quizzes to test your movie knowledge
• —Data Analysis: Get statistics on movie ratings, genres, and trends

🔧 Technical Details

• —Framework: Flask
• —AI/ML: LangChain, OpenAI, Groq
• —Vector Store: FAISS with OpenAI embeddings
• —Vision: BLIP model for poster analysis
• —Deployment: Docker on Hugging Face Spaces

📖 Examples

Movie Search:

User: "Show me action movies from 2020"
Agent: [Lists action movies from 2020]

Quiz:

User: "Let's play a director quiz"
Agent: [Starts interactive quiz about movie directors]

Poster Analysis:

User: [Uploads poster]
Agent: "This appears to be 'The Matrix' (1999) - Sci-Fi, Action"

🔗 Links

• —GitHub Repository
• —Documentation

📝 License

MIT License - See LICENSE file for details