K O L A

Inspiration

The inspiration for this project was when I was listening to a podcast and I wanted to research the people mentioned in the project, but I didn't want to skim through the video. This situation also reminded me of the recorded lectures my teacher has done and the time I wasted trying to find the content I needed. This led me to create K O L A for students like myself to have a better experience with recorded lectures/podcasts (videos).

What it does

K O L A is a web app that uses assembly ai’s Summarization, Topic Detection, and Word Timings. This web app is made possible with the Steamlit framework and is used to find the key points of an mp3 file. With the ai word timings it detects the time stamps of the mp3 file with its topic detection and provides a summary of that said topic. It is meant for students to watch lectures more effectively and cram that last-minute study sessions.

How I built it

I used Assembly AI and python to create K O L A. First, I had to learn how AssemblyAI works and how to use it. After that, I created a script that takes an MP3 file and transcribes it into highlight and chapters of the MP3 file using AssemblyAI's audio intelligence. Lastly, I used the Streamlit framework to showcase the video and time stamps AssemblyAI has converted into a JSON file.

Challenges I ran into

The greatest challenge was learning AssemblyAI. It was an advanced AI that took me by surprise. It had a wide range of features that were complex and drained my confidence.

Accomplishments that I'm proud of

A great accomplishment I have done was submitting a project by myself. I am attending this hackathon virtually, alone, and as a beginner hacker. Completing a project such as K O L A, was only the beginning leaving me to crave more.

What I learned

I learned a vast amount of things during this hackathon. I learned how to use Streamlit to showcase my application. But most especially, I learned AssemblyAI. I watched tutorials, read the documentation, and even tried out the real-time speech-to-text. I believe that learning AssemblyAI is a big thing since it can help those around me with mental disabilities. Learning AssemblyAI was only the beginning to learn so much more.

What's next for K O L A

There is so much more for K O L A. One thing I would implement is a feature where you can just submit a link to the video so that it can transcribe it without having to download it as an mp3 file. However, the most important thing I would like K O L A to have is a more interactive website.

Built With

• assemblyai
• python
• streamlit