Aurora

Inspiration

Our application was born out of the recognition that documents often contain a wealth of information, but accessing specific details can be a cumbersome and time-consuming task. To simplify this process, we developed a user-friendly solution. By allowing users to upload their documents, our application, powered by our intelligent assistant Aurora, can swiftly provide accurate answers to specific questions based on the document's content. With this automated approach, we eliminate the need for laborious manual page chunking, making it effortless for users to extract valuable insights from their documents.

What it does

Our application serves the purpose of receiving one or more documents provided by the user and generating answers to their questions based on the context derived from the document(s). By leveraging the content within the document(s), our application provides relevant and accurate responses to the user's inquiries. The goal is to offer a seamless experience where users can extract valuable information and insights from their documents through interactive question and answer functionality.

How we built it

We utilized a combination of HTML, CSS, and JavaScript to create an user-friendly front-end interface. For the back-end implementation, we opted for the Python Flask framework, which provides a robust foundation for handling server-side operations. To enhance the natural language processing capabilities, we integrated LangChain and harnessed the power of the OpenAI GPT-3 API.

Challenges we ran into

Developing this application presented our team with several challenges. Our team consists of developers with diverse skill sets, and some members had never worked with Python Flask for website development. However, we recognized that Python is the ideal language for interacting with OpenAI APIs. Determined to overcome these obstacles, we dedicated ourselves to learning Python and were ultimately successful in getting the application up and running. It was a valuable experience that allowed us to acquire new skills and accomplish our goals.

Accomplishments that we're proud of

We are able to create a platform that would help users to save time and boost their productivity. Most of the people are busy and any application that helps them saves their time is a great tool. We are proud that we are able to integrate machine learning into otherwise, a bare bone web application and make it more interactive. Since, more than half of our team are beginners we learnt a lot from this project.

What we learned

We learned about Python flask and how to interact with the large language models like ChatGPT and HuggingFace. We also learnt how to work as a team and make a successful project.

What's next for Aurora

In the future we would like to add speech-to-text and text-to-speech features for visually impaired people which would help them to interact with the application.

Built With

• chatgpt
• css3
• flask
• html5
• javascript
• langchain
• openai
• python