An intelligent Retrieval-Augmented Generation (RAG) chatbot for answering FAQs using domain-specific datasets. The bot indexes FAQs into Qdrant Cloud (vector DB) and retrieves relevant context to generate accurate answers with Groq LLM.
Built with modular Python code, this project demonstrates end-to-end RAG pipelines β from embeddings to vector search to response generation.
- π Semantic Search: Retrieves relevant FAQ context from Qdrant Cloud
- π Domain-Specific Indexing: Import datasets (CSV/HuggingFace) and index easily
- β‘ Groq-powered LLM: Fast and concise answer generation
- π οΈ Local & Cloud Support: Choose between in-memory DB (quick test) or Qdrant Cloud
- π₯οΈ Notebook Demo: End-to-end workflow in Jupyter Notebook
- πΈ Screenshots included for quick visualization
- Python π
- SentenceTransformers (MiniLM embeddings)
- Qdrant Cloud (vector DB)
- Groq API (answer generation)
- Pandas, NumPy
- Jupyter Notebook for demo
| Indexing Data | Query Answering | Multi-Query Demo |
|---|---|---|
| Indexing | Single Query | Multi Query |
All screenshots are stored in the
assets/folder.
git clone https://github.com/nv2105/QuickRAG-FAQ-Bot.git
cd QuickRAG-FAQ-Botpip install -r requirements.txt
Create a .env file in the project root:
QDRANT_URL=your_qdrant_cluster_url
QDRANT_API_KEY=your_qdrant_api_key
GROQ_API_KEY=your_groq_api_key # optional but recommendedjupyter notebook QuickRAG_Demo.ipynb
QuickRAG-FAQ-Bot/ βββ assets/ # Screenshots & diagrams βββ data/ # FAQ CSV datasets βββ src/ β βββ rag.py # Main RAG pipeline β βββ embeddings.py # Embedding generator β βββ db_qdrant.py # Qdrant Cloud integration β βββ db_local.py # In-memory DB (testing) β βββ load_data.py # Indexing script β βββ config.py # Env key loader β βββ __init__.py βββ QuickRAG_Demo.ipynb # Jupyter Notebook demo βββ requirements.txt βββ README.md
Final Year CSE Student | Aspiring Data Analyst π« LinkedInβ β’ GitHubβ
This project is open source and available under the MIT License.β