Welcome to the Complete AI Recommendation System! 🎌

You now have everything you need to run your own AI-powered anime recommendation system.

📦 What's Included

✅ Complete Source Code

• RAG Pipeline (`robust_rag_pipeline_rars.py`)
• FastAPI Backend (`api_recommendation_backend.py`)
• Streamlit Frontend (`streamlit_recommendation_frontend.py`)
• Configuration Files (`.env`, requirements.txt, etc.)

✅ Complete Dataset (8,248 anime)

• Parquet format with all metadata
• Pre-processed for RAG system
• Ready to load and use

✅ 8,248 Anime Images

• Poster Images (vertical format, optimized)
• Backdrop Images (horizontal format for slider view)
• High-quality files
• Total size: ~2.2GB

✅ Pre-Built FAISS Index

• Vector embeddings already computed
• No need to generate embeddings yourself
• Instant deployment - just load and run
• Saves hours of processing time

✅ Documentation

• Setup guides (Windows, Mac, Linux)
• API documentation
• Troubleshooting guide

💻 System Requirements

Minimum Specs (Lightweight Mode):

• RAM: 4GB+
• CPU: 4+ Cores
• Storage: 5GB+ (for models, dataset, and images)
• OS: Windows 10/11, macOS (M1/M2 supported), or Linux (Ubuntu 20.04+)
• Python: 3.10 or higher

Recommended Specs (Full AI Power):

• RAM: 8GB+
• GPU: NVIDIA GPU (4GB+ VRAM) or Apple Silicon (M1/M2/M3)
• Storage: 8GB+ (to store the generative LLM)

⚠️ Note on Performance:

The system automatically detects your hardware:

• With GPU: Enables HyDE (Generative AI) for smarter intent translation (e.g., understanding "No ecchi" logic).
• CPU Only: Automatically switches to Lightweight Mode (Vector Search + Keyword Boosting) to ensure fast performance without lag.

🎯 Why Tier 2 is Worth It

⚡ Save 20+ Hours of Development

• No need to build RAG pipeline from scratch
• No need to scrape anime data
• No need to find and optimize images
• No need to design UI/UX
• Pre-configured and tested system

💰 Save Hundreds in Costs

• No API costs for embeddings (~$50-100)
• No need to hire developers (~$500-2000)
• No design costs for UI (~$200-500)
• Ready-to-deploy solution

🚀 Production-Ready Code

• Clean, documented, maintainable
• Error handling and logging
• Security best practices
• Scalable architecture

📸 Screenshots

🖼️ Beautiful UI Design

🎨 Grid View - Apple-Inspired Cards

• Gorgeous card design with smooth animations
• Color-coded similarity badges
• Hover for instant info overlay
• 18+ content warnings

🎬 Slider View - Immersive Experience

• Full-screen backdrop images
• Glassmorphism design
• Thumbnail navigation strip
• Keyboard shortcuts (← →)

🎯 Smart Filters & Controls

• Dynamic filter updates (no re-search needed!)
• Year range: 1917-2025
• Rating threshold: 0-10
• Anime type selector
• Real-time API statistics

🔍 Similar Anime Discovery

• Find anime like your favorites
• Recursive exploration
• Color-coded match percentages
• Instant results (<0.1s)

🚀 Quick Start Guide

1️⃣ Install Dependencies

# Create virtual environment
python -m venv testenv
source testenv/bin/activate  # On Windows: testenv\Scripts\Activate

# Install packages
pip install -r requirements.txt

2️⃣ Start Backend API

uvicorn AiMi_Recommendation_System:app --reload --port 8000 <- PORT_ADDRESS

3️⃣ Start Frontend

streamlit run AiMi_Recommendation_System/frontend/streamlit_recommendation_frontend.py

4️⃣ Open Browser

Navigate to: http://localhost:8501

That's it! Your AI anime recommendation system is live! 🎉

💡 Features Breakdown

🔍 Semantic Search

• Natural language queries ("dark fantasy with strong female lead")
• 768-dimensional vector embeddings
• FAISS-powered similarity search
• Sub-2 second response times

🎲 Similar Anime Discovery

• Find anime similar to any title
• Pre-computed embeddings for instant results
• Recursive exploration (find similar to similar)
• 85-95% similarity accuracy

⚡ Smart Caching

• 6-hour query cache
• 60%+ cache hit rate
• <0.5s for repeat queries
• Automatic cache management

🎨 Dual View Modes

• Grid View: Quick browsing, 3-column layout
• Slider View: Immersive experience with arrow keys
• Switch anytime with one click
• Fully responsive (mobile-friendly)

🎯 Dynamic Filters

• Year range (1917-2025)
• Minimum rating (0-10)
• Anime type (TV/Movie/OVA/and more...)
• Results update automatically when filters change

📊 Real-Time Stats

• Total searches processed
• Cache hit rate percentage
• Average response time
• API health status

🛠️ Tech Stack

Backend

• FastAPI - Modern async Python API
• Sentence Transformers - Nomic v1.5 embeddings
• FAISS - Vector similarity search
• Pandas - Data processing
• DiskCache - Query caching

Frontend

• Streamlit - Beautiful Python web framework
• Custom CSS - Apple-inspired design
• Responsive Layout - Works on all devices
• Interactive Components - Sliders, dropdowns, cards

AI/ML

• Nomic v1.5 - 768-dimensional embeddings
• RAG Architecture - Retrieval-Augmented Generation
• Cosine Similarity - Matching algorithm
• Semantic Search - Meaning-based queries

📁 Project Structure

aimi-rag-system
├── AiMi_Recommendation_System/
│    ├── backend/
│    │   ├── __init__.py                           # Marks AiMi_Recommendation_System as a Python package and exposes core app objects.
│    │   ├── api_recommendation_backend.py         # FastAPI server
│    │   └── robust_rag_pipeline_rars.py           # RAG pipeline
│    ├── frontend/
│    │   ├── streamlit_recommendation_frontend.py  # UI
│    │   └── splash_screen_loading.html            # Intro animation
│    ├── __init__.py                               # Backend package for AiMi. Contains FastAPI API and RAG pipeline modules.
│    └──.env                                       # config file
├── assets/
│    ├── fonts/
│    │   ├── Macondo                               # font family
│    │   ├── Nunito                                # font family
│    │   ├── Poppins                               # font family
│    │   ├── Roboto                                # font family
│    │   └── Syne                                  # font family
│    ├── screenshots/
│    │   ├── screenshot1.png                       # png image
│    │   ├── screenshot2.png                       # png image
│    │   ├── screenshot3.png                       # png image
│    │   ├── screenshot4.png                       # png image
│    │   └── and so on...                          # png image
│    ├── static/
│    │   ├── 404_page.html                         # 404 html file
│    │   ├── image1.png                            # png image
│    │   ├── image2.png                            # png image
│    │   └── image3.png                            # png image
├── Dataset/
│    ├── anime_index/
│    │   ├── anime_dataset_nomic.parquet           # Complete parquet dataset file
│    │   ├── faiss_nomic.index                     # faiss index
│    │   └── metadata_nomic.json                   # model metadata
│    ├── Backdrop Images/
│    │   ├── image1.png                            # png image
│    │   ├── image2.png                            # png image
│    │   ├── image3.png                            # png image
│    │   └── and so on...                          # png image
│    ├── Logo Images/
│    │   ├── image1.png                            # png image
│    │   ├── image2.png                            # png image
│    │   ├── image3.png                            # png image
│    │   └── and so on...                          # png image
│    ├── Poster Images/
│    │   ├── image1.png                            # png image
│    │   ├── image2.png                            # png image
│    │   ├── image3.png                            # png image
│    │   └── and so on...                          # png image
├── AiMi_UI_Demo/                                  # directory including demo UI videos and .webp files
│    ├── intro.mp4                                 # loading screen
│    ├── UI_clip1.mp4                              # AiMi UI Demo 
│    ├── UI_clip2.mp4                              # AiMi UI Demo
│    ├── intro_output.webp                         # loading screen
│    ├── UI_clip1_output.webp                      # AiMi UI Demo 
│    └── UI_clip2_output.webp                      # AiMi UI Demo
├── INSTALLATION_GUIDE.md                          # Comprehensive Guide
├── license.md                                     # Must read license
├── thank_you_card1_tier2.png                      # Token of appreciation (Thank You Card)
├── thank_you_card2_tier2.png                      # Token of appreciation (Thank You Card)
└── requirements.txt                               # Containing all important libraries

🆘 Quick Troubleshooting

| Issue | Solution |
|-------|----------|
| Port already in use | Change port number in command |
| FAISS index not found | Ensure full Tier 2/3 package extracted |
| Streamlit not starting | Run `pip install streamlit --upgrade` |
| Model download fails | Check internet connection |

🌍 Deployment Options

Local Development

• Perfect for testing and customization
• Runs on Windows, Mac, Linux
• No cloud costs

Cloud Deployment

• Heroku - Free tier available
• Railway - Easy deployment
• AWS/GCP/Azure - Production-scale
• Vercel - Frontend hosting

🆙 Upgrade to Ultimate (Tier 3)

Want the Anime Receipts Generator too?

Front Side (Receipts):

Back Side (Receipts):

👑 Tier 3 - Ultimate ($149.99)

Everything in Tier 2, PLUS:

✅ Anime Receipts Generator (full application)

✅ Create Unlimited Anime Receipts with your imagination.

✅ Commercial use license (sell receipts!)

Save $50 by upgrading now!

Upgrade to Tier 3

💬 Resources

📚 Documentation

• Complete setup guides for all platforms
• API reference with examples
• Deployment tutorials
• Troubleshooting FAQ

🔒 License

🔑 License & Rights

✅ What You CAN Do (Make Money):

• Build a SaaS: Deploy a recommendation site/app and charge users.
• Freelance: Build projects for clients (deploy the app for them).
• Sell Outputs: Sell the receipts, API access, or recommendations.
• White-label: Remove AiMi branding and use your own logos.
• Deploy: Host on any server (AWS, Vercel, DigitalOcean).

❌ What You CANNOT Do (Piracy):

• Resell the Code: You cannot sell the raw source code/zip file itself.
• Open Source: You cannot upload the code to public GitHub/Kaggle.
• Redistribute Data: You cannot sell the raw parquet/index files separately.

See license.md for specific details on Client Work and Asset Usage.

Contact

For licensing questions, commercial usage inquiries, or copyright notices:

[aimi.anime.system@gmail.com](mailto:aimi.anime.system@gmail.com)

Note: For installation help or technical issues, please refer to the documentation and troubleshooting guides included in your download package.

🎉 Thank You

You now have a robust production-ready AI recommendation system.

What's Next?

1. ✅ Complete the 5-minute setup
2. 🎨 Customize the UI to match your brand
3. 🚀 Deploy to the cloud
4. 📢 Share your version with the world!

I can't wait to see what you build! 🌟

💬 Made with ❤️

Created by Divyanshu Singh - Passionate Programmer & Die Hard Anime Fan

Version 1.0 | Last Updated: 2025

Your success is my success. Now go build something amazing! 🎌

"Whatever you can do, or dream you can, begin it. Boldness has genius, power and magic in it." — Johann Wolfgang von Goethe