Context-Aware Chatbot Using RAG

A conversational chatbot that remembers context and retrieves external information using Retrieval-Augmented Generation (RAG) with LangChain, ChromaDB, local BGE embeddings, and Google Gemini 2.5 Flash.

Features

• Context-aware conversations with chat history
• Document retrieval from a vectorized knowledge base
• Local embeddings — no API needed for document indexing
• Support for PDF and TXT documents
• Deployed with Streamlit UI

Setup

1. Install dependencies:

   pip install -r requirements.txt

2. Download the local embedding model (one-time):

   python saving_embedding_model.py

   This creates the local_bge_model/ folder. Embeddings run entirely offline after this.

3. Create a .env file with your Google API key:

   GOOGLE_API_KEY=your_api_key_here
   

   Get your API key from Google AI Studio.

4. Add documents to the knowledge_base/ folder (PDF or TXT format).

5. Run the app:

   streamlit run app.py

6. Open your browser to the URL shown in the terminal (usually http://localhost:8501).

Project Structure

├── app.py                      # Streamlit UI
├── document_processor.py       # Document loading and chunking
├── vector_store.py             # ChromaDB vector store with local BGE embeddings
├── rag_chain.py                # RAG chain with Gemini 2.5 Flash
├── saving_embedding_model.py   # One-time script to download the embedding model
├── knowledge_base/             # Your documents here
├── local_bge_model/            # Locally saved BGE embedding model
├── chroma_db/                  # Auto-generated vector store
├── requirements.txt
└── .env

How It Works

1. Documents are loaded and split into chunks
2. Chunks are converted to embeddings using a local BGE model and stored in ChromaDB
3. When you ask a question, the chatbot finds the most relevant chunks and sends them to Gemini 2.5 Flash along with your conversation history
4. Gemini generates an answer based on your documents

See DOCS.md for a detailed developer guide.