Aurora

Inspiration

We developed our application after realizing that documents hold a vast amount of valuable information, but finding specific details can be tedious and time-consuming. To address this challenge, we created a user-friendly solution. With our intelligent assistant, Aurora, users can effortlessly upload their documents and receive precise answers to specific questions based on the content. This automated approach eliminates the need for manual page chunking, making it seamless for users to extract valuable insights from their documents.

What it does

Our application aims to streamline the process of receiving and analyzing user-provided documents to generate accurate answers to their questions. By utilizing the context from the documents, our application delivers relevant and precise responses to the user's inquiries. Through interactive question and answer functionality, we offer a seamless experience, allowing users to extract valuable information and insights from their documents with ease.

How we built it

Our user-friendly front-end interface was developed using HTML, CSS, and JavaScript. For the back-end, we leveraged the Python Flask framework, ensuring a robust foundation for server-side operations. To bolster our natural language processing capabilities, we integrated LangChain and harnessed the power of the OpenAI GPT-3 API. This combination enabled us to deliver a seamless and powerful application experience.

Challenges we ran into

Creating this application posed numerous challenges for our diverse team of developers. Some members had little experience with Python Flask, which was essential for interacting with OpenAI APIs. Undeterred, we took on the challenge, dedicating ourselves to learning Python and overcoming obstacles. In the end, our determination paid off, as we successfully launched the application. This experience proved invaluable, as it allowed us to expand our skillset and achieve our objectives.

Accomplishments that we're proud of

Our platform offers users a time-saving and productivity-boosting solution. In a busy world, any application that streamlines tasks is highly valuable. We take pride in integrating machine learning into our otherwise simple web application, making it more interactive and user-friendly. The project has been a great learning experience, especially for our team of beginners, as we gained valuable insights throughout the development process.

What we learned

Throughout the project, we gained proficiency in Python Flask and the intricacies of interacting with large language models such as ChatGPT and HuggingFace. Additionally, the experience taught us valuable teamwork skills, enabling us to collaborate effectively and achieve project success.

What's next for Athena

Moving forward, we aim to incorporate speech-to-text and text-to-speech functionalities into the application, catering to the needs of visually impaired individuals. This addition will enable seamless interaction with the application, enhancing accessibility and usability for all users.

Built With

• react.js