Signify AI

Signify AI is a smart email client that combines advanced AI with seamless DocuSign integration to automate workflows, compose emails, and simplify document management.

Comment

Inspiration

The inspiration for Signify AI came from the everyday challenges of managing an overwhelming number of emails and documents while trying to maintain productivity. I realized the need for an intelligent tool that not only helps users stay organized but also leverages the power of AI to draft professional emails, predict workflows, and streamline document signing processes using DocuSign. The goal was to combine the efficiency of AI with practical features to transform the way people handle their communication and documentation.

What it does

  • AI-Powered Email Composition: Draft professional emails tailored to your context in seconds.
  • Smart Email Search: Use Retrieval-Augmented Generation (RAG) to search through emails and find specific information effortlessly.
  • DocuSign Integration: Automate document workflows by identifying the right templates and enabling quick signing and sending of agreements with intelligent AI agents.
  • Automated Workflows: Predict and prepare document templates for signatures based on email content, eliminating manual tasks.

How we built it

  • Frontend: Built using Next.js, with a clean and intuitive UI for email and document management.
  • Backend: Powered by LangChain for AI workflows and integrated with DocuSign’s REST API for document signing capabilities.
  • AI Models: Leveraged OpenAI for natural language processing tasks like email composition, summaries, and contextual replies.
  • RAG (Retrieval-Augmented Generation): Implemented using LangChain to enhance search capabilities across emails and documents.
  • Database: Used Prisma and PostgreSQL to efficiently manage user data, emails, and document templates.
  • Email Retrieval: Used the Aurinko API to retrieve and sync emails from the users' Inbox. I am incredibly grateful to the Aurinko team for providing me with an extended trial usage on their API so I can build this project! This project would not have been possible without this API.
  • ** Custom Built APIs:** In addition to this, I built approximately eight internal APIs to manage document workflows, user authentication, database read/write/update operations, email and inbox management and also APIs to leverage the power of OpeanAI and Langchain to aid in the retrieval augmented generation process.

Challenges we ran into

  • DocuSign Integration: Understanding and implementing DocuSign’s API workflows, especially for templates and dynamic roles, was a complex task.
  • Streaming AI Responses: Implementing smooth streaming of AI responses in real-time posed technical challenges.
  • Inbox Search with RAG: Ensuring accurate and efficient email search using RAG required fine-tuning embeddings and vector-based search algorithms.
  • User Authentication: Managing secure user authentication and session handling with Clerk while ensuring a seamless experience.

Accomplishments that we're proud of

  • Successfully integrating DocuSign’s API to automate workflows and enable seamless document management directly within the email client.
  • Implementing advanced AI capabilities like email composition, RAG-powered search, and contextual replies.
  • Building a polished, user-friendly application that merges communication and documentation tasks into one platform.
  • Solving real-world challenges like reducing time spent on email management and simplifying document signing processes.

What we learned

  • The power of combining multiple tools like LangChain, OpenAI, and DocuSign to create a cohesive and intelligent workflow.
  • How to optimize AI-powered applications for performance and accuracy, especially in search and contextual tasks.
  • Insights into user behaviour and the potential for automating repetitive email and document-related workflows.

What's next for Signify AI

  • Enhanced Analytics: Provide users with actionable insights about their email and document trends.
  • AI Agent Evolution: Develop more intelligent agents capable of deeper workflow automation, such as identifying potential bottlenecks or providing proactive suggestions.

  • Template Management: Develop the template management feature more to allow for the creation of new templates and all types of relevant documentation to be handled from Signify AI.

  • Please note that once the hackathon ends the provided test link will not work due to trial versions of APIs etc coming to an end.

Built With

Share this project:

Updates