EzNotes

• 

• 

  Home page

• 

  Generated notes

Inspiration

In today's world students are looking for ways which save their time as they have to balance time between their studies and social life. So, we aim to create a product which helps in doing smart work – Being efficient and doing more work in a certain amount of time. We do this by generating AI-powered lecture notes for students since when a student tries to watch a lecture and take notes at the same time, typically neither process is done effectively, therefore by allowing the student to spend more time on understanding rather than recording, we are able to simplify learning.

What it does

Our platform uses OpenAI and machine-learning-powered solutions to summarize YouTube videos and uploaded media into concise and enriching bullet points. student would input a youtube link in the input bar, the youtube video would then be analyzed and key points would be extracted from the video transcript. the points extracted from the transcript would finally be presented in bullet points. The student would also have the option to download a word document in order to save their notes locally or share it with others.

How we built it

We built EzNotes using the following technologies:

Next.js: We used Next.js in order to build the frontend and render all the pages. Next.js was used because it enables makes page rendering efficient and quick, and the developer experience of Next.js is very good and allowed our team to be productive, and effectively build the most optimized frontend possible.

Flask: Flask is a lightweight backend framework in Python. We used flask to build the backend API for our application. Since it was lightweight, we felt that Flask would be ideal for this usecase and time limitations. Moreover, flask is a relatively performant framework to other web framework alternatives for Python.

GPT-3: In order to summarise a YouTube video lecture into brief bullet points, we felt that the best solution would be to use the largest text generational platform ever. Therefore, we used OpenAI's GPT-3 to extract key points from the video transcript and generate notes as bullet points.

Challenges we ran into

The first challenge we faced was trying to find a field where we as students needed improvement, This required multiple hours of research and brainstorming possible solutions as well as trying to create something unique. The other challenge we faced was trying to implement the GPT-3 API and Open AI into our program. This was our first time using GPT-3 hence running into any issues regarding this took significantly longer.

Accomplishments that we're proud of

Our team is proud of creating a working product that can help students save time and work more efficiently in the time limit provided for the hackathon. Being able to create a fully functional product within the time span provided showed the effect of our time management skills also shows our skill level as a developer. The ability to incorporate the GPT-3 API into a program and using it for the first time expands our limit into becoming a better developer.

What we learned

Throughout the process of creating EzNotes, We learnt about the issues faced by students in regard to time management as well as working with GPT-3 and Flask. This was our first time working with flask and we were able to use it to its fullest extent. GPT-3 usage provided an insight into using it for further projects to help the student community further.

What's next for EzNotes

In the future, we hope to expand EzNote's capabilities by building our own audio transcription model which would students to upload their own lecture videos which would be transcribed in real-time and then the notes would be generated. We would also like to add the ability to use videos from other video streaming platforms like Dailymotion or Vimeo, rather than simply YouTube.

Built With

• flask
• gpt3
• nextjs
• python