The project involves creating a Python-based AI chatbot that uses the OpenAI API to handle two core scenarios for our business: customer support and personalised product recommendations. All interaction will happen through a text-based web chat interface, so the focus is on natural-language understanding, role-based responses, and a clean user experience rather than voice or mobile features. I would like to collaborate with a developer located elsewhere in Asia (Pakistan, India, China, or the Philippines are ideal) so that our working hours overlap and communication stays quick. Budget: $200 What I need delivered • A well-structured Python script (preferably 3.10+) that calls the OpenAI API, separates “support” and “recommendation” roles, and logs each session. • A quick-start installation guide that lets me spin the bot up on a fresh Linux machine with minimal commands. Virtual-env, requirements.txt and any environment-variable setup should be included. • A short screencast video (or GIF) that walks through the running demo, showing both support-style and recommendation-style conversations. • Inline comments and concise docstrings so the codebase is immediately readable. Acceptance criteria 1. Chatbot answers at least 20 common support questions from a provided FAQ file with >90 % accuracy in manual testing. 2. Recommends products from a supplied JSON catalogue based on user intent and returns top three items ranked by relevance. 3. All external secrets (e.g., OpenAI key) are loaded from .env and never hard-coded. 4. Setup guide can reproduce the working bot on my side within 15 minutes. If you have links or repos from similar OpenAI or GPT-powered chat projects, please include them so I can gauge fit quickly. I’m ready to get started as soon as the right approach is clear and should be available for daily check-ins during Asian business hours.