Project Info
Inspiration
Teachers spend countless hours preparing lesson plans, but one thing is difficult to practice before entering the classroom... the classroom itself. A lesson rarely goes exactly as planned. Students ask unexpected questions, lose focus, become disengaged, or require different levels of support. While AI has made it easier to generate lesson content, there are still very few tools that allow teachers to rehearse the actual teaching experience. As someone passionate about building AI products that solve real-world problems, I wanted to explore how AI could empower educators rather than replace them. That led to ClassPilot, an AI-powered classroom simulation where teachers can practice lessons, build confidence, and receive coaching before teaching real students.
What it does
ClassPilot is an AI classroom simulator that allows teachers to rehearse lessons in a realistic virtual environment. Teachers simply upload a lesson plan, choose classroom settings and student profiles, then begin a live teaching session. During the simulation, AI students behave independently by: Asking spontaneous questions Raising their hands Taking notes Becoming bored or distracted Responding differently based on their personalities A live class engagement meter reflects how well the class is responding, encouraging teachers to adapt their teaching style in real time. At the end of each session, ClassPilot provides coaching insights to help teachers improve future lessons.
How we built it
I built ClassPilot using Next.js and Tailwind CSS for the frontend, while leveraging Gemini Live API for low-latency real-time conversations. The core of the project is an event-driven classroom simulation engine. Each AI student is assigned a unique personality and behaviour profile, allowing the classroom to react dynamically instead of following scripted conversations. The simulation continuously generates classroom events such as questions, disengagement, note-taking, or raised hands while tracking overall class engagement throughout the lesson.
Challenges we ran into
The biggest challenge was making the classroom feel natural rather than scripted. Building a conversational AI is one thing, but simulating an entire classroom where multiple students behave independently required designing an event engine that balances realism without overwhelming the teacher. Another challenge was managing real-time interactions while keeping the experience responsive. I wanted conversations to feel fluid, with classroom events happening naturally alongside live voice interactions.
Accomplishments we're proud of
Built a functioning AI classroom simulation within a single week Designed an event-driven classroom engine that creates dynamic student behaviours Integrated real-time voice conversations with AI students Developed a live classroom engagement system to simulate teaching effectiveness Created a product that focuses on supporting teachers through practice rather than replacing them Most importantly, I am proud of turning an educational idea into an interactive experience that teachers can genuinely learn from.
What we learned
This project reinforced that great AI products are about experiences, not just models. I learned how to combine real-time AI conversations, event-driven simulations, and thoughtful interaction design to create something that feels immersive and useful. I also discovered that building believable AI behaviour requires careful orchestration, not just prompting an LLM, but designing systems that simulate realistic human interactions.
What's next
I'm excited to continue evolving ClassPilot beyond the hackathon. Our roadmap includes: Enhance our AI classrooms experience that closely mirror real teaching environments Expand student personalities, learning styles, and classroom dynamics to better reflect real-world teaching scenarios Multi-language support Introduce session playback and reflection tools Integration with lesson planning workflows used by educators My long-term vision is simple: Help every teacher teach with confidence by practicing before they step into the classroom.
This project did not link a GitHub repository.
Analysis
No indexed repository for this project, so there are no commit stats to show.
Technology
- Google GeminiUnchecked
- Next.jsUnchecked
- SupabaseUnchecked
No repository was indexed for this project, so these Devpost claims have not been checked against code.
AI coding agents
No repository was indexed, so agent usage could not be checked.
Detected from committed agent config files and commit authorship. Absence of a signal is not proof an agent was unused.
Codebase size
No repository was indexed, so there is no codebase to measure.
This project did not link a GitHub repository, so there is nothing to diagram.
This project did not link a GitHub repository, so its feature claims have not been checked against code.
Export this project's context (description, README, evidence, key source files) to chat with an AI agent elsewhere.