PublicaX

Inspiration

PublicaX started from a very simple idea: people see problems in their city every day, but reporting them is usually slow, confusing, or ineffective. Things like potholes, blocked drains, broken streetlights, or unsafe roads often stay unfixed not because no one cares, but because the information doesn’t reach the right people in the right way. We wanted to explore how technology and AI could make that process easier and faster.

While working on PublicaX, we learned a lot about building for two completely different types of users. On one side are citizens, who just want a quick and easy way to report an issue. On the other side are city staff, who need clear, organized, and prioritized information to do their jobs efficiently. Designing a system that works well for both was one of the biggest lessons from this project.

PublicaX was built as a web-based prototype. Citizens can report issues using photos, videos, text, or voice notes, with the location captured automatically. The system then uses simple AI logic to categorize the issue and estimate how urgent it is. City staff see everything in a separate dashboard, where reports are listed and shown on a map to make patterns and critical areas easy to spot.

One of the main challenges was deciding what to include in the MVP and what to leave out. We also faced challenges around handling media uploads, separating user access, and keeping the AI simple but meaningful. Overall, this project helped us understand how AI can support real-world public services without overcomplicating them.

Built With

• fastapi
• gemini
• next.js
• postgresql
• python
• supabase