📘 AI Pokédex – RAG-Powered Pokémon Assistant

An AI-powered Pokédex that lets users explore Pokémon information using natural language queries, powered by Retrieval-Augmented Generation (RAG) and a local LLM.

This project is fully CPU-friendly, does not require a GPU, and runs completely offline using Ollama.

---

🚀 Screenshots

🖥️ ChatGPT-Style Pokédex UI

---

📋 Pokédex Card View

---

🔄 Pokémon Comparison

---

🧠 What This Project Does

• 🔍 Answer Pokémon queries in natural language
• 🧠 Uses RAG (FAISS + embeddings) for factual accuracy
• 🤖 Generates descriptions using a local LLM (Ollama)
• 🖼️ Displays Pokémon images
• 📊 Shows Pokédex-style stat cards
• 🔄 Supports Pokémon comparisons (e.g., Dialga vs Palkia)

---

✨ Key Features

✅ Natural Language Chat

Ask things like:

• Tell me about Pikachu
• What type is Solgaleo?
• Dialga vs Palkia

---

🧠 Retrieval-Augmented Generation (RAG)

• FAISS vector search
• Sentence-Transformer embeddings
• Ensures answers are grounded in Pokémon data

---

📊 Pokédex Cards

Each Pokémon card includes:

• Image
• Type badges
• Height & weight
• Abilities
• Base stats
• AI-generated description

---

🔄 Pokémon Comparison

• Side-by-side Pokémon analysis
• Stats & type comparison
• Ideal for battle insights

---

💻 CPU-Only & Offline

• No GPU required
• Runs locally using Ollama
• Optimized for low-end systems

---

🏗️ Architecture Overview

        User
         ↓
  Streamlit Chat UI
         ↓
  FastAPI Backend
         ↓
 RAG Engine (FAISS)
         ↓
 Pokémon Dataset (CSV)
         ↓
  Local LLM (Ollama)

---

⚙️ Installation & Setup

1️⃣ Clone Repository

git clone https://github.com/<your-username>/ai-pokedex.git
cd ai-pokedex

🗂️ Project Structure

ai-pokedex/
├── backend/
│ └── app/
│ ├── services/
│ │ ├── rag_engine.py
│ │ ├── pokedex_parser.py
│ │ ├── name_extractor.py
│ │ ├── image_resolver.py
│ │ └── llm.py
│ └── main.py
├── frontend/
│ └── streamlit_app.py
├── data/
│ ├── pokemon.csv
│ └── images/
├── scripts/
├── screenshots/
│ ├── chat_ui.png
│ ├── pokedex_card.png
│ └── comparison.png
├── requirements.txt
├── .gitignore
└── README.md

---
Create Virtual Environment
python -m venv venv
venv\Scripts\activate

Install Dependencies
pip install -r requirements.txt

Install Ollama & Model

Download Ollama from https://ollama.com

ollama pull phi3
# or
ollama pull llama3:8b

Run Backend
uvicorn backend.app.main:app --reload

Run Frontend
streamlit run frontend/streamlit_app.py


Open http://localhost:8501

---

🧪 Example Queries

• hello
• Tell me about Abra
• Pikachu vs Raichu
• What are Solgaleo's abilities?

---

🧰 Tech Stack

• Python
• FastAPI
• Streamlit
• FAISS
• SentenceTransformers
• Ollama (Local LLM)
• Pandas
• NumPy

---

👨‍💻 Author

• Ankush Maity
• B.Tech CSE
• AI / ML Enthusiast
• GitHub: @Amsir390

---

🌟 Why This Project Stands Out

• Real-world RAG implementation

• Local LLM (no API cost)

• Industry-ready architecture

• Strong portfolio and interview project

---

⭐ Support

• If you like this project, give it a star on GitHub!