Inspiration: Tutorials teach concepts, but I wanted students to practice the kind of work they would face on the job.
What it does: Students create a fictional company, explore a 3D office, and choose one of five AI coworkers. They can learn a topic, complete a realistic task, ask questions, submit their work, and improve it using feedback. Activities include coding, written assignments, and customer conversations.
How I built it: I used Next.js, React Three Fiber, Monaco, and Meta Muse Spark. Redis, Milvus, Neo4j, and SQLite support memory and source-based learning.
Challenges and lessons: The hardest part was making AI-generated tasks feel grounded in a real workplace. I learned that the most useful coaching gives students something to do, checks their work, and helps them try again.
Built With
- amazon-web-services
- langraph
- python
Log in or sign up for Devpost to join the conversation.