AskMe – AI-Powered Document Chat Assistant

AskMe is an intelligent document-based chatbot that allows users to upload files (PDF,TXT) and instantly ask questions about their contents.
It combines RAG (Retrieval-Augmented Generation) with Ollama LLMs for accurate, context-aware answers — running completely locally for privacy and speed.

---

🚀 Features

• ⚡ RAG-based QA – Retrieve contextually relevant answers
• 🧠 Local LLM Integration – Uses Ollama (e.g., Llama 3) for inference
• 🔍 Efficient Vector Search – Powered by FAISS
• 💬 Interactive Chat UI – Conversational interface for real-time queries
• 🪶 Lightweight & Fast – Optimized chunking and embeddings

---

🧰 Prerequisites

Before running AskMe, ensure the following are installed:

• Python 3.10+
• pip (latest version)
• Ollama → https://ollama.ai
• Git

Then, pull the required models:

ollama pull llama3
ollama pull nomic-embed-text

---

⚙️ Installation

# Clone the repository
git clone https://github.com/<your-username>/AskMe.git
cd AskMe

# Install dependencies
pip install -r requirements.txt

---

▶️ Run the Backend

uvicorn main:app --reload

---

💻 Frontend (Optional)

If your project includes a frontend UI:

npm install
npm run dev

---

⚡ Tech Stack

Component	Technology
Backend	FastAPI
LLM Engine	Ollama (Llama 3)
Embeddings	nomic-embed-text
Vector Store	FAISS
Framework	LangChain
Frontend (optional)	React + Tailwind

---

🧩 How It Works

1. User uploads documents
2. Text is split into chunks and embedded
3. Chunks are stored in a FAISS vector database
4. When a user asks a question:
   • Relevant chunks are retrieved
   • Query + context are sent to Llama 3
   • The model generates an accurate answer

---

💡 Author

Sanjay Muthuswamy – AI Enthusiast & Innovator

---