RAG-Based AI Assistant – Module 1 Project

This project implements a Retrieval-Augmented Generation (RAG) AI assistant, built as my solution for the Agentic AI Developer Certification, Module 1.

Overview

The assistant leverages the RAG architecture to answer open-ended questions using my own uploaded knowledge base of .txt documents. The system uses document chunking, embeddings (all-MiniLM-L6-v2), ChromaDB for search, and integrates with multiple LLM APIs (Google Gemini, Groq, OpenAI). It replies accurately using only the uploaded context, so answers are trustworthy and verifiable.

Key Implementation Steps

• Loaded custom .txt files from the data/ folder
• Split documents into chunks for semantic search
• Embedded these chunks and stored/query in ChromaDB
• Designed a prompt to use retrieved context, answer, or say 'I don't know' if the info wasn't in my docs
• Built/run a question-answer pipeline using LangChain and tested with multiple topics
• API/LLM used: [Gemini/Groq]

Usage

• Clone the GitHub repo: https://github.com/wasxy47/rt-aaidc-project1-template
• See README for setup and running instructions

Notes

• No API keys or sensitive data are stored in the public repo
• Add your .txt files to customize your assistant’s domain easily