AI Student Task OrganizerDuely
Founded and led a three-person team building an Android-first student organizer that turns assignment photos into editable tasks through on-device OCR and optional AI assistance.
← Back Open to Opportunities
Wave & Wish
Led a six-person team to deliver a touchless festival game in seven days, serving 280+ players with zero crashes.
/CASE STUDY
The challenge
Build a camera-controlled festival game that could survive a seven-hour public event, difficult lighting, rapid player turnover, and a fixed physical-prize inventory.
Technical approach
- Mapped MediaPipe palm landmarks to screen coordinates with exponential smoothing, a deadzone, and position prediction for brief tracking loss.
- Designed a dynamic difficulty system, anti-streak spawn rules, three fully themed content sets, and an event-driven mascot HUD.
- Built an auditable gacha system with configurable win rates, multiple wishes, unique claim codes, and JSONL result logging.
- Packaged the Python application for Windows with a custom PyInstaller specification covering MediaPipe models and OpenCV dependencies.
Results
Event-ready game delivered in seven days
280+ players served in seven hours
170 physical prizes distributed
0 crashes during the event
