Inspiration

I've been working on creating products using publicly available Large Language Models (LLMs). When Dropbox announced this hackathon, I was captivated by the idea of combining the capabilities of LLMs with the Dropbox Sign API.

What it does

The app empowers users to effortlessly generate documents by providing a simple description. They can engage in a conversation with the AI to provide finer details, ensuring the generation of a more refined document. Once the document is generated, users can seamlessly edit it within the app using our integrated document editor. After they are satisfied with their edits, they can proceed to save the document. Subsequently, users can send the document and request signatures. The app allows for the issuance of multiple signatures for the same document and provides real-time monitoring of the signature status. Users can send reminder emails and even cancel signature requests if necessary. This process can be repeated for multiple documents.

How we built it

The foundation of our document generation lies in the GPT-4 chat completion API. Through proper prompt engineering, I have harnessed its power to generate text documents. These documents are then seamlessly passed to the in-app word editor. Upon user save, we convert the document into a PDF format. When a user requests a signature, we transmit the file along with the collected details to Dropbox Sign's "Send Signature Request" API. To retrieve the status of the request, we utilize the "Get Signature Request" API. For sending reminder emails, we employ the "Send Request Reminder" API.

Accomplishments that we're proud of

I embarked on this project as a last-minute endeavor and managed to construct the entire solution within a single day. It's truly astonishing to witness how an application facilitating the generation and signing of legal documents can materialize so swiftly, thanks to the remarkable capabilities of publicly available LLMs and the powerful APIs of Dropbox Sign.

What's next for AI Doc Gen

Our future plans involve refining the product further and potentially releasing it to the public as a fully-fledged solution.

Built With

Share this project:

Updates