Legal-Eaze

Empowering the everyday person with the understanding of legalese through machine learning.

Comment

Inspiration

Legal documents can be daunting and difficult to understand for the average person. Our project aims to bridge the gap by developing a website that utilizes natural language processing to provide concise summaries of key law cases and precedents. This proof-of-concept demonstrates our ability to simplify legal jargon and make it more accessible to the general public. In the future, we hope to expand this technology to assist individuals in comprehending their personal legal documents such as lease agreements and terms-of-service agreements, enabling them to make informed decisions. Our ultimate goal is to promote legal literacy and empower individuals to navigate the complex legal landscape with confidence.

What it does

Legal-Eaze is a versatile chatbot designed to simplify legal documents. It currently supports manual input of legal text; in the future we hope to support pdf files as inputs as well. Our system utilizes a Name Entity Recognition (NER) natural language processing (NLP) model to extract relevant objects from the data. We then provide clear and concise summaries of relevant court cases, precedents, and other potentially dense legal jargon.

How we built it

We have developed a user-friendly front end for our chatbot using React. Our homepage and chat interface are designed to provide users with an intuitive and efficient experience. For classification of input data, we utilized an open-source NLP model and dataset. To summarize legal topics, we leveraged OpenAI's API and models. We implemented Convex, a TypeScript-native programmable database for the web, to enable seamless transfer of data from our front end written in JavaScript to our backend written in Python. Furthermore, we integrated Checkbook's API to provide a donation button and facilitate streamlined donation payments. Our chatbot is built with robust technology and an easy-to-use interface to empower users to navigate complex legal jargon with ease.

Challenges we ran into

One of the biggest hurdles we faced was the lack of accessibility to lawyer-tagged data in English. While we intended to incorporate US legal data, the most detailed dataset we found was specific to India's legal system. However, we persisted and were able to utilize this dataset to create a proof-of-concept for our chatbot. Another challenge we faced was integrating the convex database, a technology that was new to our team. Despite these obstacles, we overcame them and successfully implemented the convex database, enabling seamless transfer of data between our front end and back end. Our team is proud of the progress we've made and we look forward to continuing to improve our chatbot.

Accomplishments that we're proud of

One of our key achievements was the seamless integration of multiple frameworks. The ability to integrate multiple APIs and provide straightforward responses to users is a testament to our team's dedication and expertise. We are confident that our chatbot will be a valuable resource for individuals seeking to navigate the complexities of legal documents with ease.

What we learned

Our team consisted of members with varying levels of experience in web development, and working on this project was an excellent learning opportunity for all of us. We were able to expand our skillsets and gain exposure to a wide range of new technologies, including the creation of a React app and utilizing web server frameworks with Flask. In addition, we successfully integrated APIs from emerging technologies such as OpenAI's API and Checkbook's API. This project challenged us to think creatively and adapt to new technologies quickly, ultimately allowing us to enhance our abilities and grow as developers.

What's next for Legal-Eaze

Our chatbot has the potential to be trained to comprehend and simplify a broader range of legal topics and documents. Initially, we plan to focus on improving the chatbot's understanding of personal documents such as terms of service agreements and housing leases. Achieving this will require us to train the chatbot on more diverse and complex legal language and to expand its natural language processing capabilities. We also aspire to train the chatbot to understand and simplify legal language in other languages, which would make it more accessible to non-English speaking users. This would involve incorporating language-specific models and datasets to enable the chatbot to understand legal jargon in a variety of languages. Our team recognizes the potential for further development of our chatbot and is committed to enhancing its capabilities in order to provide users with an invaluable resource for navigating the complexities of legal documents.

Share this project:

Updates