Apartlo

Inspiration

As a college student myself, finding housing and discovering all of the hidden fees of purchasing an apartment was a very frustrating experience, to say the least. I often asked myself, "Why is this process so difficult?" The answer is simply that companies don't want the process to be easy. That's why I made Apartlo. Fast, easy, and simplified searching for students renting their first apartment.

What it does

What we aim to do is simplify the apartment search and reveal hidden details for students searching for housing. Some things that are commonly hidden on front pages are security deposits, application fees, required insurance purchases, and utilities. In addition to revealing hidden fees, Apartlo also allows users to prompt our website's AI to auto-apply filters to make the process of finding apartments faster for users. The AI also creates a summary of why the apartment listings it found are relevant to users and provides a 0-100 rating of why that apartment matches them. Using AI tools to help empower students and their search for housing can create a positive impact on students and relieve the headache of doing the research themselves.

How we built it

Using Google Gemini and Antigravity IDE, detailed prompts were provided to create robust and sophisticated user flows, user interfaces, and API connections within the Apartlo project. APIs used for this project include Gemini 2.5 for our AI model, Tavily for AI connections to the web, and Supabase for Apartlo's database.

Challenges we ran into

This project was created on a heavy time crunch, and due to this, a lot of user flows were not fully checked to account for edge cases, and features such as Google Authentication could not be implemented. AI isn't perfect either, meaning that many of the prompts had to be detailed and robust, which took up a majority of the time creating the project as well.

Accomplishments that we're proud of

The number of APIs and the fact that the product is fully functional are astonishing in the amount of time that it was created. I'm happy with what I created and look forward to seeing what I can implement with it in the future.

What we learned

I learned how to provide refined, detailed, and robust AI prompts to effectively create a product that's functional. Often times I find myself not covering certain areas that AI might miss, and AI is only as strong as your prompt (as most would say). Being able to improve on my prompts and connecting my knowledge of full-stack development in such a short amount of time really taught me a lot about how to effectively create something feasible.

What's next for Apartlo

There are many edge cases, user flow errors, and UI improvements that can be made, and I want to be able to polish that in the future. There are more features I want to add as well, such as Google Authentication, but unfortunately, that was not something feasible in the time that I had. In the future, I want to be able to improve on this project as I'm passionate about student housing and simplifying the process of apartment searches.

Built With

• gemini
• node.js
• react
• supabase
• tavily
• typescript
• vercel
• vite