Hall of Works The Trial , Case study
Lantern of Focus : LUMIO
AI-powered early attention detection and learning support. Tracks student focus in real time and gives evidence-grounded guidance.
- Role
- Co-founder, AI and full-stack
- When
- 2025
Vanquished The Drifting Mind 1st place, Sight Day Congress Technical Challenge
- MediaPipe WASM
- XGBoost
- Ensemble models
- LangChain
- FAISS
- RAG
- n8n
- JavaScript
- Python
The Foe , The problem
Early attention difficulties are often noticed late, after a student has already fallen behind. Teachers and parents need signals that are early, objective, and explained, not a black-box score.
The Arms , Approach and stack
- In-browser vision: MediaPipe WASM tracks face and gaze signals on the student’s device. No video leaves the machine.
- Classification: XGBoost and ensemble models turn those signals into focus / distraction states.
- Guidance: three AI assistants built on a LangChain + FAISS RAG engine, grounded in evidence.
- Orchestration: every workflow runs through n8n, so it is transparent and auditable.
Victory , Results
- 1st place, Sight Day Congress Technical Challenge (LUMIO and cognitive diseases).
