YText | Devpost

The landing page

Inspiration

Our group is composed mostly of beginners, and as such we have never made used Assembly AI before. Although YouTube does provide transcription on some videos, they are hard to navigate and hard to find the information you are looking for. A lot of us take notes based off YouTube videos daily, so we thought that making a website that does the work for us would help make things more convenient for us and our peers.

What it does

YText is a website, which has a homepage with FAQ, About Us, and a place to enter a Youtube video URL. It uses Assembly AI to transcribe any YouTube video to create a readable page. This allows for students and organizations to extract information off videos easily as they do not have to sit through the entire video and wait for the required information, but rather can save time by using our website.

How we built it

We built the front end using HTML, CSS and frameworks such as bootstrap. The backend was built using Python as that was what most of us were most comfortable using. We used AssemblyAI as the AI will transcribe any audio recording to speech using the speech to text feature.

Challenges we ran into

While building, we ran into many challenges. Domain.com had a backend crash while did not allow us to register our ytext.tech domain. AssemblyAI was new to all of us so reading the documentation and trying to understand how to use it was a major challenge.

Accomplishments that we're proud of

Although we ran into many challenges, we are very proud of the end product. From never using the AssemblyAI to managing to create a well-functioning and designed application that can be implemented into daily use, we are proud of ourselves. This was also the first time we worked with each other, so it was an interesting experience learning with each other and having fun while doing so.

What we learned

We learned how to use AssemblyAI to convert speech to text.

What's next for YText

We plan on using the new AssemblyAI summarizing speech feature to create chapter based notes off the YouTube video. This would completely revolutionize the note taking process as it would be fully automated. Additionally, we plan to implement additional features in the websites home screen and add more advanced features such as shortcuts and modals.