This project started with an attempt to perform web scraping on Dubizzle Egypt for real estate data. However, due to the way the company's website was built, it was challenging to extract clear and usable data.
After encountering issues with Dubizzle, Bilal collaborated with Nassar to collect real estate data from the Aqarmap website. This alternative source provided the necessary data more effectively, allowing us to proceed with the project.
Once the data was gathered, Bilal and Ali started working on developing the user interface using Streamlit. This allowed for a simple and interactive frontend for users to engage with the application.
The frontend was connected to an API-powered LLama model using the Hugging Face model hub. The application generates real estate-related text using Llama,providing users with dynamically generated content.
The project was uploaded to GitHub, and GitHub Actions was implemented to ensure continuous integration and deployment. This setup allows the application to automatically update whenever changes are pushed to the repository, ensuring the live version is always in sync with the latest code.
The final version of the application was deployed on Streamlit, making it publicly accessible for users to interact with.
[Visit the Application]