🏠 Window Sales Chatbot

👋 AI-powered chatbot for window/door sales industry. It handles initial customer contact before they reach a salesperson.

Why this project?

I want to help salespeople in window industry by:

• 24/7 availability for customers
• Reducing workload of sales staff
• Automating measurement appointments
• Providing instant product information
• AI-powered natural conversations

Tech Stack

• Python 3.12 - Backend programming language
• FastAPI - REST API framework
• Uvicorn - ASGI server
• OpenAI GPT-4o-mini - Artificial intelligence
• HTML/CSS/JavaScript - Frontend interface
• python-dotenv - Environment variables management

Progress

Stage 1 - Basic Chatbot (COMPLETED)

• Python environment setup
• Project structure
• intents.json with 6 categories
• Pattern matching logic
• Console version (chatbot.py)

Stage 2 - REST API (COMPLETED)

• FastAPI integration
• Endpoints: /, /chat, /info, /clear
• Auto documentation at /docs
• CORS enabled for frontend
• Session-based conversation memory

Stage 3 - Frontend (COMPLETED)

• HTML chat interface
• JavaScript API integration
• Real-time bot responses
• Clickable links support
• New conversation button

Stage 4 - OpenAI Integration (COMPLETED)

• GPT-4o-mini model connected
• Natural language understanding
• Context-aware responses

Stage 5 - Advanced RAG (COMPLETED)

• Company knowledge base (wafam_oferta.txt)
• Text chunking (56 fragments)
• OpenAI embeddings (text-embedding-3-small)
• ChromaDB vector database
• Semantic search
• Conversation memory
• Topic tracking
• Optimized token usage

Stage 6 - To Do

• Conversation logging
• Lead collection
• Admin panel
• Deployment to cloud

How to run?

Step 1: Clone repository

git clone https://github.com/Kajhol/window-sales-chatbot.git
cd window-sales-chatbot

Step 2: Create virtual environment

python -m venv venv

Step 3: Activate virtual environment

Windows CMD:

Set-ExecutionPolicy -ExecutionPolicy RemoteSigned -Scope CurrentUser
venv\Scripts\activate

Step 4: Install dependencies

pip install fastapi uvicorn openai python-dotenv langchain langchain-openai langchain-community chromadb

Step 5: Create .env file

Create file .env in window-sales-chatbot folder: OPENAI_API_KEY=your-api-key-here
Get your API key from: https://platform.openai.com/api-keys

Step 6: Build knowledge base (first time only)

cd window-sales-chatbot/src  
python prepare_knowledge.py  
python build_vectordb.py

Step 7: Run API server

cd window-sales-chatbot/src
uvicorn api:app --reload

Step 8: Open frontend

Open in browser: window-sales-chatbot/frontend/index.html

📫 Contact

Questions or code review? Find me here:

• LinkedIn: https://www.linkedin.com/in/kajetan-hołdan-9b4a503a0/
• Author: Kajhol (Computer Science Student, Silesian University of Technology, 3rd year)

---

Star this repo to follow progress!