KAWAMART/DIUMART

DiuMart Ghana is an AI-powered e-commerce website developed for a CS305 Artificial Intelligence project. It features product search, categories, shopping cart functionality, personalized customer assistance, and an NLP-powered chatbot that answers questions about products, delivery, payments, returns, refunds, policies and orders. # DiuMart Ghana — AI E-Commerce Chatbot **CS305 Final Project — AI-powered e-commerce website and chatbot** DiuMart Ghana is a Streamlit-based e-commerce website with an integrated AI shopping assistant. It demonstrates three core chatbot requirements: 1. **The right communication channel** — the assistant is embedded directly on the homepage (not a separate page), so help is available the moment a customer arrives. 2. **Effective feedback/error messages** — when the chatbot can't confidently answer, it gives a clear fallback message and escalates to human support instead of guessing. 3. **Personalized communication** — the chatbot collects the customer's name, favourite category, and budget, and uses them to tailor product recommendations and replies. ## Features - Product catalogue of ~1,000 items across 10 categories, with search, category filter, and price filter - Shopping cart and checkout with mock order creation - AI chatbot (rule/keyword-based) that can: - Greet customers automatically - Capture and use personalization details (name, favourite category, budget) - Recommend products matching a query, category, and/or budget - Answer FAQs (delivery, payment, returns, refunds, order tracking, cancellations, warranty, privacy, security/PIN safety) - Add items to the cart via chat - Fall back gracefully and hand off to human support for anything it can't answer confidently - FAQ, Policies, Order Tracking, and Human Support sections - Security policy explicitly stating DiuMart will never ask for a customer's Mobile Money PIN, banking password, or OTP ## Project structure ``` DiuMart/ ├── app.py # Main Streamlit application (site + chatbot) ├── products.csv # Product database (~1,000 products) ├── requirements.txt # Python dependencies └── README.md # This file ``` ## Running locally ```bash pip install -r requirements.txt streamlit run app.py ``` The app will open at `localhost`. ## Deploying to Streamlit Commun …