Ignorance is the root cause of all that is evil in today's housing market. We personally realized that only after signing a bad lease. That inspired to help people like us who sign up for things without actually realizing the full extent of how it can affect their lives, and in a way to improve the housing inequality due to lack of knowledge.

What it does

Lease genie essentially acts as another set of eyes that goes through your lease document thoroughly and highlights the key points that you should note before signing a lease for a house. And that's not all! Considering to invest in a house. We have all the stats you need to consider under one roof!

How we built it

Built with the aim to reduce the housing inequality, we created a Flask web app, which is a portal to check the facts of documents being uploaded and also gives a detailed insight on the tax and buying patterns of a neighborhood. The web app leverages Google Cloud products AI powerhouse to scan through a document and validate it with law points. Also, the Natural Language is heavily applied over the document to extract the key entity-relationship. In addition, we were able to provide some insights on the buying pattern in the neighborhood the user wishes to buy a house, deriving data from Zillow and the presented the tax pattern additionally.

Challenges we ran into

The number of housing documents is very sparse and we were not able to identify many. This caused an issue with the ML model which required data to learn. Also, housing data was not public and Zillow data had its own limitations. We had to learn Flask, though we thoroughly enjoyed it.

Accomplishments that we're proud of

We were able to implement the Redaction system and bring in data-privacy. Also, we're able to go through the Tenants rights and understand many fundamental rules, which formed the crux of this project. We were glad to do data analytics over the Zillow data to find insights on the buying pattern in places. The central document processing and suggestion providing portal have rich AI combined, to provide results in a short period, with detailed Natural language processing involved.

What we learned

The accuracy of Google Cloud ML model was good, but it still needs lot of housing data to learn more. We learned using Google Cloud APIs and web scraping! We were amazed to see the number of housing laws of NY being little represented in many leases.

What's next for Lease Genie

With sufficient training data, we are hoping to extend this feature to buying/selling contracts and this could be an excellent add-on for ibuying feature recently introduced in Zillow

Built With

Share this project: