Inspiration
We were inspired by this project by becoming aware of the fact that people do not read contracts before signing it. We encountered a CBS News article showing that nearly 70% of people do not read their contracts. Life is short, and people do not have all day to read the extensive legalese of a contract, along with the 20 pages worth of clauses. Therefore, people pick up the pen before even reading the contract, which can lead to problems, including:
- accepting unfavorable terms, which can lead to disadvantages later for you/your company
- hidden fees and costs, which are commonly buried deep inside the contract
What it does
Our app tackles exactly that. We have a file upload system, where one can upload a file of the contract that needs to be analyzed. The app then processes the text, and prompts the user with the text hat is going to be analyzed. After that, the Gemini API jumps in, helping analyze the legal document, along the following points:
- clause title
- clause number
- risk score
- issues
- who it benefits
- ambiguities
- a concise yet comprehensive summary Finally, the app returns the data as a table, with each of the points organized within corresponding rows. There are also filters available, which allows for users to zoom in on the key points.
How we built it
We built this program using HTML, CSS, and JavaScript. We used Node.js, and the Gemini API in order to do the backend work.
Challenges we ran into
We ran into many challenges, including the configuration of the Gemini API. Since I am new, I struggled quite a lot with this step, and I had to look at many tutorials in order to understand it. Also, because this app was tested extensively, we frequently surpassed the limit for the amount of queries that I could make. However, I learned a lot
Accomplishments that we're proud of
We are proud that we finally finished our first AI powered project. I also practiced and increased my skills of Node.js web servers, Git, etc.
What we learned
We learned how to configure node.js into my application. We also learned how to configure the Gemini API. Finally, but most importantly, I learned to persevere when times are tough. I had encountered many bugs in my code, but I learned that I must not give up.
What's next for Docalyzer
-Convert it to an app that can be posted to Google Play or the App Store -Accessibility tools -OCR character recognition for physical documents -Improve UI/UX
Log in or sign up for Devpost to join the conversation.