Overview
SeekLiyab is a web-based fire detection system built for electrical engineering students at PUP who needed a working IoT prototype. Sensor data flows into a Random Forest classifier that flags fire conditions in real time, hitting 89% detection accuracy. The Streamlit interface shows live sensor readings and triggers alerts when thresholds are breached. Supabase handles the backend — storing sensor logs and supporting the real-time data pipeline. The goal was a system that could actually be deployed, not just a demo.
Commissioned by Electrical Engineering students at Polytechnic University of the Philippines.
Gallery
