I’m building a public-facing website where investors can swap proven tips, and I want the experience to feel truly intelligent rather than just another bulletin board. The core feature is an AI engine that learns from user-supplied data and surfaces personalised investment recommendations each time someone logs in. Think of it as a tailored “ideas feed” that updates as the member’s watch-list or risk profile evolves. Here is what I already have in mind: a clean, mobile-friendly front end (React or similar is fine) that hosts discussion threads, article pages, and a simple onboarding flow collecting the risk tolerance and preferred sectors of every new member. Your job is to wire that to a back-end model—whether you favour Python with FastAPI, Node, or another stack—to generate those recommendations in near real time. You’re free to plug in an API such as OpenAI or build a light custom model if you prefer; accuracy and explainability matter more to me than the specific library. In addition to coding the recommendation logic, I’ll need: • A small administrative dashboard where I can review the model’s suggestion history and flag any odd results. • Clear documentation so that future contributors understand the architecture, dataset format, and update routine. • A quick hand-off call after deployment. Security (OAuth, encrypted storage for any sensitive user details) and a straightforward CI/CD workflow are expected as part of the delivery. If you’ve shipped similar fintech or data-driven platforms, let’s discuss your approach and timeline.