What web and AI project could I realistically complete for my final-year project?

A practical option would be a project-idea recommendation website. Users could enter their interests, skills, available time, and preferred technologies, then receive suitable project suggestions. The web portion could handle accounts, forms, search, and saved ideas, while a simple AI model could classify users’ requirements or recommend projects from a prepared database. It solves a genuine problem and can be expanded gradually without making the core version too difficult.

Start by discussing the requirements and your current skill level with your mentor. They can help you narrow the scope and choose technologies that are realistic for a year-long project. A small, complete system is much better than an ambitious idea that never gets finished.

Choose something with a basic version you can finish early, then add AI features only after the main application works. For example, build a study-planning or resource-recommendation website, using AI to categorize resources or generate a draft plan. Get regular feedback from your mentor, keep the scope controlled, and focus on learning the fundamentals rather than trying to create a large commercial product.