Inspiration

We are students who sometimes don't manage our time correctly. Commuting to class makes it a bit difficult to study on the go. That's why we made Pod-Studious. We sometimes end up seeing articles that we'd be pretty interested in reading, or topics that we might not understand too well before a big test but trying to read while driving is dangerous. We wanted a free and easy way to be able to read articles or generate a short form podcast about a specific topic using LLMs.

What it does

The application has 3 mains functions the first one is a web scraper that can read the main body of an article to you. The second one generates a podcast using a specific topic that you input. The third and final function is a conversation AI function, where users can conversationally learn, asking questions to the AI agent.

How we built it

We built this with a React frontend and a Python + Fast API Server for the Backend. Gemini LLM for our prompting and our Generate Podcast and Converse AI pages.

Challenges we ran into

We ran into numerous challenges, we were having issues connecting our frontend and backend. We had issues containerizing our project, as well as setting up a database as we were all unfamiliar with this. We opted to not use a database because we were not storing any information long term. We wanted to develop and train our own LLM but didn't work out. We are also trying to get text to speech working so that we are able to "converse" with our AI.

Accomplishments that we're proud of

We are proud of being able to have functionality in our app. We were able to get both the Article Reader and the Podcast Generator working perfectly fine. Our UI also looks pretty cool in our opinions!

What we learned

We learned how to finish a hackathon! But we learned a lot like making LLM API calls and working with their

What's next for Pod-Studious

We are going to continue to work on the Conversation AI and try our best to get speech to text implemented so that we can have continuous conversations. We also would like our audio to have the personality of celebrity's and mimic their voice and conversation style.

Built With

Share this project:

Updates