AI Discovery Prototype for Social Experiences Mobile App

I have an existing mobile app. The app helps users find other people to attend attractions, events, and experiences together. I want to test a new AI discovery prototype. Instead of users only browsing a normal ticket catalogue, the user should be able to type something like: “I’m free Saturday in London, under £30, and I don’t want to go alone.” The AI should then suggest suitable experiences or crowds based on: mood / intent location budget date and time social preference existing or sample crowd data This is not a full rebuild and not a full production release. I want a small working prototype first. The prototype should include: AI prompt/input screen concept recommendation output with 3–5 suggested crowds or experiences simple matching logic explanation of why each crowd is recommended options such as Join Crowd, Create Crowd, and Already Booked backend/API approach that can later integrate into the existing app use of OpenAI API or similar LLM provider clear notes on how to avoid the AI inventing fake prices, events, or availability Important: the AI should recommend from provided/sample data, not hallucinate real attractions, prices, or availability. Deliverable can be either: clickable prototype, simple frontend demo, backend API demo, or staging-ready logic that my current developers can review. Please include examples of any AI apps, recommendation systems, chatbots, travel/event apps, or mobile app integrations you have built. Have you integrated OpenAI API into a mobile app or web app before? 1. How would you structure the recommendation logic? 2. How would you stop the AI from inventing fake results? 3. Can you deliver a small prototype within 7–14 days? 4. What exactly would be included in your fixed-price proposal? The goal is to test whether the app can become an AI-powered social experience discovery app, where users say what they feel like doing and the app helps them find a crowd.