🧠 Technical Interview RAG Chatbot

This is a full-stack Retrieval-Augmented Generation (RAG) chatbot system designed to help users prepare for software engineering technical interviews. It combines local document retrieval with a large language model (LLM) to answer questions based on custom educational content like PDFs, slides, spreadsheets, and more.

---

🚀 Features

• Multimodal document ingestion: PDF, DOCX, PPTX, CSV
• Local embedding using sentence-transformers
• Vector store using ChromaDB
• Conversational retrieval with contextual prompt building
• FastAPI backend API
• Streamlit frontend with modern chat UI
• Modular Python architecture (easy to extend)

---

⚙️ Setup Guide

Follow the steps below to install dependencies, process documents, and launch the chatbot:

1. Clone the repository

git clone https://github.com/StevenGX12/RAG_Chatbot.git
cd technical-interview-rag

2. Set up the virtual environment

python -m venv myenv
source myenv/bin/activate       # On Windows: myenv\Scripts\activate

3. Install dependencies

pip install -r requirements.txt

4. Start Ollama (if using a local model)

Ensure you've installed Ollama and pulled a model:

ollama run gemma3

5. Add documents to the corpus

Place all your learning materials under the folder:

./corpus/

Supported formats: .pdf, .docx, .pptx, .csv

6. Run the document processing pipeline

python -m src.scripts.update_db

This will:

• Parse and chunk new documents
• Generate embeddings for unembedded chunks
• Add them to the ChromaDB vector store

Start the FastAPI backend

uvicorn src.app_backend:app --reload

Runs on: http://localhost:8000

Launch the Streamlit frontend

In a separate terminal:

streamlit run src.app_frontend.py

File Structure

src/
├── scripts/                  # Utilities: update_db, backfill
├── app_backend.py           # FastAPI server
├── app_frontend.py          # Streamlit chat UI
├── data_loader.py           # Corpus parser and processor
├── embedder.py              # Embeds document chunks
├── retriever.py             # Query-time document retrieval
├── vector_store.py          # ChromaDB logic
├── generator.py             # Prompt building and LLM calls
processed_corpus/            # Output: chunks and embeddings
corpus/                      # Input: source documents
assets/                      # UI assets like screenshots