Sim.ula

The story

In Quezon City, traffic congestion costs about 3.5 billion pesos a day, and some areas have up to 40,000 people per square kilometer. In a city like that, a bad policy decision is extremely costly, and the government can't afford to experiment on its own streets. We wanted to give planners a safe place to test ideas before putting them into practice.

During a 24-hour hackathon, my team and I built sim.ula, an AI platform for simulating urban policy. It draws on a database of 94 real policies from cities across Southeast Asia, such as Singapore and Jakarta, and checks how well each one would actually transfer to Quezon City. It also maps how a single change can ripple out into long-term effects. Since public policy is high-stakes, we added a separate model that checks the AI's recommendations so it can't present made-up information as fact. The project won 3rd place overall. I built it alongside Charles Joshua Uy and Prince Rivera.