A web-based AI chatbot that answers professional questions using a knowledge base with fuzzy string matching. Built with React frontend and FastAPI backend.

• Hf Spaces Deployed Link: https://huggingface.co/spaces/lydiaavin/mini-ai-chatbot

• Frontend (React + Vite): Clean conversational UI with input field, send button, answer display, and chat history.
• Backend (FastAPI): REST API with /ask endpoint for question processing.
• Knowledge Base: 56 professional Q&A pairs covering productivity, leadership, finance, marketing, HR, and more.
• Matching: Fuzzy string matching using fuzzywuzzy library with 50% similarity threshold.
• History: Persists chat history in history.json for context in responses.
• Fallback: Provides a professional fallback message for unmatched questions.

• Python 3.8+ (for backend)
• Node.js 14+ (for frontend)

1. Backend:

   cd backend
   pip install -r requirements.txt
   uvicorn app:app --reload --host 0.0.0.0 --port 8000

2. Frontend (in a new terminal):

   cd frontend
   npm install
   npm run dev

3. Open http://localhost:5173 in your browser to access the chatbot UI.

• Endpoint: POST /ask
• Request Body: {"question": "Your question here"}
• Response: {"answer": "Answer text"}

curl -X POST http://localhost:8000/ask \
  -H "Content-Type: application/json" \
  -d '{"question": "How to manage time effectively?"}'

Response: {"answer": "Set SMART goals, use calendars, and avoid multitasking."}

• Backend: FastAPI, Uvicorn, Pydantic, fuzzywuzzy
• Frontend: React, Axios, Vite

mini-ai-chatbot/
├── backend/
│   ├── app.py                 
│   ├── knowledge_base.json    # ~50 Q&A pairs
│   ├── history.json           
│   └── requirements.txt       
├── frontend/
│   ├── src/
│   │   ├── App.jsx            
│   │   ├── App.css            
│   │   └── main.jsx           
│   ├── package.json          
│   └── vite.config.js         
├── .gitignore
└── README.md

• Knowledge base covers common professional topics; unmatched questions receive a fallback response.
• Fuzzy matching threshold of 50% balances accuracy and coverage.