Lawyers spend a lot of time reading legal documents and also many people cannot understand/ extract important terms from a legal contract and hence, it is a waste of money and time just to read contracts. Our motivation is to build software that will save lawyers time and people time while reading the contract
What it does
It allows users to upload contract documents and get answers to legal questions (binary type) and a contract summary in just a few clicks.
How we built it
We used pandas, NumPy for data processing, and GPT 3 transformer model, and also for frontend we used React with GPT 3 - API
Challenges we ran into
Shortage of time and sleep deprivation.
Accomplishments that we're proud of
Our model is working with decent accuracy and our UI/UX is roughly done
What we learned
This is the first time we build the whole model pipeline with API so we have learned a lot regarding the machine learning model and GPT3- APT
What's next for ContractXTract
We will create a feature for pdf file reader and host our website for users to try and update/retrain our model pipeline for better accuracy
Log in or sign up for Devpost to join the conversation.