Module 1 – RAG Q&A Assistant
Overview

This project is a Retrieval-Augmented Generation (RAG) based Q&A Assistant designed to answer user questions by retrieving relevant information from a set of local documents.
It demonstrates how AI can combine semantic search with natural language understanding to deliver accurate and context-aware answers.

How It Works

Document Loading – Text files stored in the data/ folder are automatically loaded and converted into vector embeddings.
Embedding Generation – The project uses the sentence-transformers/all-MiniLM-L6-v2 model to transform text into semantic vectors.
Vector Search – The embedded data is indexed in a ChromaDB vector database, allowing efficient retrieval of relevant context.
Answer Generation – The context is passed to the Groq Llama 3.1 (8B) model, which generates a final natural-language answer.
Interactive CLI – Users can type questions directly in the terminal to interact with the assistant.

Dataset

The project uses a local dataset stored in the ./data/ directory.
Each document is a plain text file containing educational or informational content used for context retrieval.
No external or online dataset was used.

Features

Retrieval-Augmented Generation (RAG) pipeline
Local document processing
Vector database integration
Fast and context-aware question answering
Modular Python architecture for learning and experimentation

#Tech Stack

Python 3.8+
ChromaDB – Vector database
SentenceTransformers – Embedding model
Groq Llama 3.1 – Large Language Model
LangChain – RAG orchestration

Run the app

python app.py

Example Interaction:

Enter a question or 'quit' to exit: What is Artificial Intelligence?
Searching for: What is Artificial Intelligence?
Model Response: Artificial Intelligence (AI) is the simulation of human intelligence in machines...

Results

The assistant accurately retrieves relevant document context and provides coherent, human-like answers — even when the input phrasing changes.
This project demonstrates how combining retrieval and generation improves the factual grounding of language models.

Author

Mitiku Tadesse
Software Engineering Student | Passionate about AI, Machine Learning, and Backend Systems