We operate across multiple platforms:
PAI Box: Our AI modeling platform. We utilize python scripts to download data, and subsequently, train three housing price prediction models based on this data. And Currently, we have a front-end interface (Python+HTML, deployed on AWS) to display the prediction results of these models.Retbox: Our operational platform that assists our team. Our colleagues inspect and manage properties, each with varying statuses. The main functionalities include workflow optimization, automation (e.g., personalized email generation based on templates), and efficient management and review of diverse property statuses. Note that Retbox imports foundational property data from PAI Box, which includes housing price prediction details for analytical purposes. Currently, a third party that collaborates with us is developing this website (using a front-end framework, which will also be deployed on AWS and requires regular maintenance).
Future Platforms: We’re looking to integrate tools like PowerBI—a potent business intelligence platform. For instance, upon uploading our data (CSV format), PowerBI could depict our data using geographical coordinates across different map layers. We’re also exploring platforms that utilize IoT sensor data and platforms oriented towards tenant data collection. In the future, We might directly purchase services from third-party companies and access them via API or other methods. Alternatively, we could develop our own applications with similar functionalities and integrate them into our platform.
Design a unified system from the perspectives of real estate and business logic. In this system or platform, we should be able to access all of the aforementioned functions and services. The purpose of this system designs is to empower our business, property analysis, and all aspects of property management with technology. Therefore, it should be logical, user-friendly, and easy to interact with.
(Below are some factors to consider. If you have more detailed considerations in the system design, that would be a plus.)
1.Please take into account scalability (for instance, as time progresses, the data volume for our AI model will grow, and consideration is needed on how to store and manage this data on AWS).
2.Please consider flexibility; many functions, including the integration of different platforms, may change based on our evolving needs.
3.Please consider customization requirements; all platforms, including many functions, need to be tailored based on our specific business requirements.
4.Please consider the timeline, cost, and efficiency of the solution.
5.Please focus on data collection and management. For instance, we internally acquire or own some data, like from Paibox. Other data might come from third parties, in varying data types and formats. We hope to aggregate and manage these tables as uniformly as possible, coupled with a sensible data management scheme. When we have business requirements, we should be able to quickly retrieve all possible information about a property through SQL for our analysis and processing.
6.We want to deploy our AI model, our database, out platform on AWS, Please describe in as much detail and specificity as possible your plan for deploying relevant applications and models on AWS.
7.Please consider the relationships between data and the data flow. For instance, the basic data for properties in Retbox comes from Paibox, and we can also import data from Paibox into PowerBI for analysis.
8.Please consider integration solutions between different platforms, API design, and so on
Please describe your thoughts and design proposal in as much detail as possible. You can illustrate your system design approach using various methods such as sketches, ER diagrams, textual descriptions, and so on.