AI Projects for High School Students

AI is no longer something you study after school — it's something you build in school. Whether you're taking your first computer science class or already training models in your spare time, there's a project on this list for you. These 16 ideas are scoped for high school students who want to build something real, learn something transferable, and end up with work they're proud to show.

You don't need a full engineering team. You don't need a semester of prerequisites. You need a good problem, a little patience, and the right starting point. That's what the Figma Make templates in this list are for. Each one gives you a working interface to react to, refine, and make your own.

Read on to learn:

• How to pick an AI project that fits your skill level and interests
• 16 project ideas, each matched to a real Figma Make community template
• Five tips for building AI projects that stand out in competitions and applications

What is an AI project, exactly?

An AI project is any project that uses a machine learning model to process input and return a useful output. That input might be text, an image, a sensor reading, or a dataset. The output might be a prediction, a summary, a classification, or a recommendation. The point isn't the technique — it's the problem you're solving.

The projects in this list span a range of difficulty and domain. Some connect to public APIs and return results in seconds. Others involve training your own model on a custom dataset. Most fall somewhere in the middle. The templates match each project's interface needs, so you can focus on the hard part: building something that actually works.

16 AI project ideas for high school students (with templates)

1. AI image classifier

Every image classifier starts with the same question: what does the model actually need to see? Start there. Describe a simple image-recognition interface in Figma Make — something like 'an app that lets a user upload a photo and see a label and confidence score returned from an AI model.' You'll get a working UI skeleton fast. Then swap in a real API like Google Vision or Hugging Face and watch it come alive. This is one of the best first AI projects because the feedback loop is immediate: you upload a photo of a dog, you see 'Labrador, 94%.' That's the moment it clicks.

Why it works

It's visual, fast to test, and the output is instantly understandable. Students who struggle to stay engaged with abstract code tend to stick with this one because the results are tangible.

Template: Animated AI Data Flow Visualization by Aditya Kumar

2. AI voice assistant UI

Voice interfaces are everywhere, and designing one reveals a lot about how AI systems actually work. Tell Figma Make you want 'a voice assistant app with a listening state, a processing state, and a response display.' Then dig into the logic: how do you handle ambient noise? What happens when the model isn't sure? The design questions and the technical questions turn out to be the same questions, which makes this project a great bridge between UX thinking and machine learning.

Why it works

Voice projects push students to think in states and transitions, not just screens. That systems thinking transfers directly to ML model design.

Template: AI Voice Interaction App by Nessi.Huang

3. AI agency or portfolio website

Want to show what you've built? Build the site that showcases it. Type something like 'a portfolio website for a high school AI developer with a project grid, an about section, and a contact form' into Figma Make and you'll have a styled, responsive site in minutes. It's a project that's both technically satisfying and practically useful. You can use it to apply for internships, competitions, or college programs. Building your portfolio inside the same tool you used to build your projects? That's a flex.

Why it works

Portfolio sites teach students to communicate their work, not just build it. That skill compounds over time.

Template: AI Agency Website by sitenoobs ceo

4. AI-powered document editor

What if your text editor could summarize, rewrite, or explain any selected paragraph on demand? Prompt Figma Make to create 'a document editor with a sidebar AI panel that accepts a selected text block and returns a summary, a rewrite, or a translation.' Connect it to the Anthropic or OpenAI API and you've got something genuinely useful. This project is a favorite for students who want to explore large language models without wading into fine-tuning — the interface work is meaty enough to anchor the whole project.

Why it works

Editors are a classic project type that touch text manipulation, API design, and UI state all at once. The AI layer makes it feel current.

Template: AI Document Editor Design by Mitali Agrawal

5. AI-assisted task manager

There's a reason task managers are a classic project. They touch everything. Tell Figma Make you want a kanban board where each card includes an AI button that suggests a time estimate and breaks the task into subtasks. Then start breaking it. What happens when a task is vague? How does the model handle ambiguity? This is where the interesting edge cases live, and chasing them down is how you learn to build systems that actually hold up.

Why it works

Task managers force students to think about data persistence, user intent, and error states — the unsexy parts of engineering that separate good projects from great ones.

Template: Kanban Board Application by dimorin

6. Personalized learning dashboard

Every learning tool starts with the same question: what does a student actually need to see? Start there. Prompt Figma Make with something specific, like 'a student dashboard that tracks quiz scores across subjects, shows a progress chart, and surfaces an AI recommendation for what to study next.' The recommendation engine can be as simple as a rule-based prompt or as complex as a retrieval-augmented system. Either way, the project teaches you to think about data, display, and decision-making at the same time.

Why it works

Ed-tech projects resonate because students are the users. The empathy you bring as the person who's actually been through school is a genuine advantage.

Template: Student Dashboard LMS Application by Sakshi Parashar

7. AI developer portfolio site

This is the portfolio site built specifically to show off AI work. Not just a project grid, but a living site where visitors can interact with your models directly. Describe it to Figma Make: 'a data science portfolio with project cards, a live demo embed, an about page, and a contact form.' Add a section where a visitor can type a prompt and see one of your models respond. That kind of interactive showcase is rare at the high school level, and it's memorable.

Why it works

A portfolio that lets people try your work, rather than just read about it, does more work for you in a college interview than any resume line.

Template: Data Science Portfolio Website by siddhesh m

8. Computer vision monitoring dashboard

Industrial AI is one of the fastest-growing fields in machine learning, and this project puts you right in the middle of it. Build a monitoring dashboard that simulates a camera feed, processes frames through a detection model, and displays alerts when something unusual is spotted. Prompt Figma Make for 'a manufacturing monitoring dashboard with a live camera panel, a detection log, and an alert system.' Connect it to a YOLO or TensorFlow model for real detection. The result looks professional and demonstrates real-world AI application.

Why it works

Computer vision dashboards are portfolio gold. They demonstrate systems thinking, API integration, and domain knowledge all at once.

Template: VisioTrack AI Manufacturing Dashboard System by Kartik Pande

9. Data analysis and upload tool

Raw data doesn't communicate on its own. Your job is to build something that does. Prompt Figma Make to create 'a data upload dashboard where a user can drop in a CSV, select columns to analyze, and see AI-generated insights and a summary chart.' Connect it to a Python backend or a tool like Pandas AI. This is a natural fit for students interested in data science who want a front-end that matches the sophistication of their analysis work.

Why it works

Data tools teach the full stack. Students learn to think about input validation, transformation, and output formatting — not just the model in the middle.

Template: GraphQL Data Analytics File Upload Dashboard (Copy) by Daph

10. AI workflow automation builder

What if non-technical users could build their own automation workflows, just by describing them? That's what this project asks you to figure out. Describe it in Figma Make: 'a dark-theme workflow builder with draggable nodes, connectors, and an AI panel that suggests the next step based on what's already in the flow.' Students who take this project seriously end up touching graph data structures, natural language parsing, and UI interaction design all in one build.

Why it works

Automation builders are hard. That's the point. The struggle teaches more about system design than any tutorial.

Template: Dark Theme Workflow Automation UI by Bhavya Chaurasia

11. IoT sensor monitoring app

AI and IoT are converging fast, and this project sits right at the intersection. Build a flowchart-based monitoring app that pulls in simulated sensor data, runs it through an anomaly detection model, and surfaces alerts when readings fall outside expected ranges. Prompt Figma Make to create 'an IoT dashboard with a sensor list, a real-time readings panel, and an alert log.' Use Python and MQTT or a mock data generator to feed the system. It's a project that's equally at home in a science fair and a job interview.

Why it works

IoT projects teach students that AI decisions have physical consequences. That shift in thinking is valuable in any engineering context.

Template: IoT Water Level Monitoring Flowchart (Copy) by Idris Ogundele

12. AI ethics debate tool

Not every AI project needs to be a working model. Some of the most important work in this field is about understanding what these systems do and why it matters. Build an interactive debate tool that presents two positions on an AI ethics question, lets the user explore supporting arguments, and uses an AI facilitator to push back on weak reasoning. Prompt Figma Make to create 'a debate interface with two position panels, a central prompt area, and an AI response section.' This is a standout project for students interested in policy, law, or philosophy.

Why it works

Ethics projects demonstrate a kind of maturity that's hard to fake. They also tend to spark the most interesting conversations in presentations.

Template: Business Process Flow Diagram by Sandeep Kumar

13. Terminal-style AI chatbot

There's something about a command-line interface that makes an AI feel more real. Build a retro-terminal chatbot where users type natural-language commands and an LLM responds in character. Describe it in Figma Make as 'a green-on-black terminal interface where a user can type commands and receive AI-generated responses with a typewriter animation.' Connect it to an OpenAI or Anthropic API with a custom system prompt. Bonus: give the AI a persona and see how far you can push the illusion.

Why it works

Terminal UIs are back in vogue and highly distinctive in a portfolio. The character design layer teaches prompt engineering in a way that abstract examples never do.

Template: Retro CRT Terminal Website by MRZ

14. AI competition leaderboard

If your school has a robotics club, a coding competition, or any kind of STEM tournament, build the leaderboard for it. Prompt Figma Make to create 'a live competition leaderboard with team scores, a ranking table, a score submission form, and an AI commentary panel that generates a match summary after each round.' The AI commentary layer is what makes it special. Bonus: use the leaderboard at the actual competition and see how people react when the AI starts narrating their performance.

Why it works

Leaderboards are deceptively complex. Real-time updates, tie-breaking logic, and data consistency are all engineering problems in disguise.

Template: Enhanced Leaderboard Table by Paridhi

15. AI research poster generator

Science fair season hits differently when you've got an AI poster generator on your side. Build a tool that takes a structured research abstract as input and outputs a formatted academic poster with sections for methodology, findings, and conclusion. Describe it to Figma Make as 'a research poster generator where a student fills in text fields and the layout updates automatically.' Add an AI layer that rewrites jargon into plain language for the summary section. It's useful immediately and demonstrates a clear design problem worth solving.

Why it works

Tools that solve real problems for the people building them tend to be the most polished. When you're the user, you care more about getting it right.

Template: Academic Research Poster Design by Maha

16. Community issue reporting app

Here's a project that uses AI to do something that matters. Build a civic reporting app where community members can submit local issues — potholes, broken streetlights, unsafe intersections — and an AI triage layer categorizes, prioritizes, and routes each report to the right department. Describe it in Figma Make as 'a mobile issue reporting app with a submission form, a map view, a status tracker, and an admin dashboard.' This project works at every level: it demonstrates technical depth, civic awareness, and product thinking.

Why it works

Civic tech projects stand out at competitions because they're harder to dismiss. The social value is built into the brief.

Template: Civic Issue Reporting App by neetu singh

Five tips for AI projects in high school

Start with a problem worth solving

The strongest AI projects begin with a question, not a technique. 'I want to build an image classifier' is a technique. 'I want to help students with visual impairments navigate our school's website' is a problem. Problems give your project a reason to exist beyond the grade. They also make your documentation, your presentation, and your demo exponentially easier to explain.

Prototype the interface before the model

It's tempting to start with the hard part, the model, and add the interface later. Flip it. Build the front-end first in Figma Make. Figure out what the user sees, what they input, what comes back. Once you can describe the interaction precisely, the model layer becomes much easier to scope. You'll also catch design problems before they become code problems.

Use real data, even if it's small

The difference between a tutorial project and a real project is usually the data. Tutorials use clean, pre-processed datasets. Real projects use messy, incomplete, sometimes contradictory data. Find a real dataset, even a small one, that's relevant to your problem. The act of cleaning and understanding it will teach you more about AI than any lecture.

Document as you go

The most common regret at project presentation time is 'I wish I'd written down why I made that decision.' Keep a running log: what you tried, what failed, and what you learned. This documentation becomes your project narrative, and it's what judges, teachers, and interviewers actually want to hear about. Figma's comment and version history features make it easy to capture design decisions without breaking your flow.

Get someone outside your team to try it

Your project isn't done until someone who didn't build it has tried to use it. Watch them. Don't explain. Take notes on where they get stuck. That observation session will surface more improvement opportunities in 20 minutes than a week of solo testing. It's also great practice for the kind of user research that professional product teams do every day.

Start building with Figma

We hope this list gives you a project worth starting. Whatever you build, build it with real data, a clear problem, and someone in mind who'll actually use it. When you're ready, we'd love for you to try Figma.

• Prompt your way to a working AI interface with Figma Make
• Build out your prototype visually in Figma Design
• Map workflows and co-create with your team in FigJam
• Present your project with live interaction in Figma Slides
• Find your next community template in Figma Community

The next great AI project is already in your head. Let's build it together.