NoteWise

logo

Inspirations

Notetaking results in about 13% higher academic performance than not taking notes, especially when combined with cued lecturing link. However, taking notes can be a time-consuming, disheartening, error-prone process, and may cause distractions. We want to make the note-taking experience smarter, smoother, and more efficient.

What it does:

based on various sources (paper/fiction/slides/etc), NoteWise can:

generate easy-to-understand explanations.
generate well-structured mind maps with detailed notes.

How you built it

Utilized Python Flask to implement a backend HTTP server and called fine-tuned OpenAI API to generate output text.
Implemented a dynamic webpage using React.js, realizing text input & output and mind map showcase.
Deployed the web app to Google Cloud E2 instance

Challenges you ran into

We have no prior experience with OpenAI API, so we encountered difficulties managing the Open API keys, and adjusting prompts and parameters to generate ideal output text.
Negotiation is another problem as our teammates are from different backgrounds (robotics, backend, frontend, ML).

Accomplishments that you’re proud of

NoteWise is a highly scalable web application integrated with state-of-the-art Large Language Models. Currently, it is a demo, it can scale up by employing different LLMs and building additional features.

Moreover, we are proud of being able to finish the project within 24 hours and publish the website with the full functionality we expected.

What you learned

We have acquired extensive experience in coordinating both front-end and back-end functionalities. Our team has honed its skills in leveraging the OpenAI API, adeptly adjusting prompts and parameters to yield optimal output text. Additionally, we have developed a strong proficiency in deploying web applications on Google Cloud E2 instances