In a world where people often skim and hate reading long documents. We wanted to solve the problem of people signing the documents, business contracts, etc in a hurry. We want to create an error proof and efficient process of contract signing.
What it does
Our product summarizes the content of documents into a few key lines which will give an overview to the user about the contract. This allows the reader to understand the summary and main points behind the text. Then using DocuSign's API, the user can share the contract to be signed by the recipient. The recipient of contract will also get the summarized text through the API of DocuSign through his email. After that the recipient will sign the contract and the sender will receive notifications on when this is done. DocuSign's main purpose is signing, but our team was able to utilize it's other features such as end to end notification on where the document is on this signing process.
How I built it
Frontend - HTML, CSS, Bootstrap Backend - Django Web Development Framework, DocuSign API, Python. Our highly efficient backend was written in Python with 90%+ accuracy in summarizing the main points of a document. We used machine learning and a ranking scheme to create this.
Challenges I ran into
- Integrating the API and text summarizer to our website.
- Some problems with libraries while uploading the project to heroku platform for hosting the website.
- Dividing the work in team members. We were all new and met at Hack MIT.
Accomplishments that I'm proud of
- Completing the project and using DocuSign API Successfully. The process is very smooth and can be used for a legal secretary.
- Building something with real use and has an impact.
What I learned
- Using DocuSign API and learning the cool features behind it.
- Working with new people in a short span of time, the DocuSign mentors help us a ton!
- Don't give up, try until the goal is reached.
What's next for Sign-Off
- Sentimental analysis that will be transferred as subject (like Urgent, Not-Urgent, etc.)
- Partnership with DocuSign or an agreement to turn Sign-Off into a startup. We ready to ship this into the real world and want to visit the DocuSign HQ and talk merger!
- Using neural network methods like LSTM, CNN, etc for text summarization. Our algorithm is 90% efficient but we want to improve this even further!
- Blockchain technology for vast network.