Real State Market Places Recommender System

Our project consist on a recommendation system for a user who wants to rent a property.

It consists on a webapp made with streamlit, where the user can upload his own photos of the property, and we give a feedback of the property and the recommendation of where to announce/post it (for example, idealista).

For the backend, we use python and the API provided by restb.ai. Once the user has uploaded the photos (from local), we store them in an azure container to get their URL in order to finally pass them to the API.

When we call the API, we ask for four solutions which are caption, features, room type, and score. We use them because we think that those are the most useful for the project since the description and recomandation we provided are based on those paramaters of the photos that the user has uploaded.