LocAI

Inspiration

Last year, my friend and I struggled tremendously while searching for student housing. We spent countless hours across multiple websites, manually comparing rent prices, commute times, and lease terms using spreadsheets. The process was overwhelming and time-consuming. This personal pain point inspired us to create LocAI (the intelligent housing assistant we wish we had).

What it does

LocAI allows users to chat with an AI bot about their housing preferences. Users can describe their ideal accommodation in natural language, such as:

"I need a 2-bedroom apartment under $1200 within 30-minute bus ride to campus"

"Show me houses with utilities included and good lease deals"

"Find pet-friendly studios near UB with short-term leases"

The AI analyzes these requirements and returns the best matching properties from our database, highlighting good deals and optimal leases

How we built it

We built LocAI as a full-stack web application with several integrated components. The frontend provides an intuitive chat interface where users can naturally converse with our AI assistant. The backend processes these conversations and handles the business logic, connecting the user interface with our intelligence layer.

We integrated advanced natural language processing capabilities to understand user preferences and extract key search parameters from casual conversation. Our data pipeline continuously collects and processes housing information from various online sources, normalizing this data into a consistent format. All property information is stored in a structured database that enables efficient querying and matching based on user criteria.

Challenges we ran into

Buffalo map remake. Data of properties scraping is hard.

Accomplishments that we're proud of

We calculated the distance from each of the properties to the university so that We can have a detail list of data that we can use to suit the users' needs.