Study Engine

Group #: 38 - OTUcoders Group Members: Malhar Singh, Vedant Vaidya, Varun Gagwani, Johnny Liang

Inspiration

As university students, we often struggle to understand lecture contents, trying to take notes while keeping up with the professor. Recognizing this challenge, we thought of a solution that could help students like us to gain knowledge and understand content. Presenting Study Engine - AI-powered companion

What it does

• Study Engine integrates Speech-to-Text to provide real-time transcription of audio content
• Allows students to ask questions with AI-generated answers to help clarify concepts instantly
• Allows user to upload files and generate a summary of important topics
• Prepares flashcards based on uploaded file

How we built it

Our website integrates both a front-end and back-end seamlessly to provide users with intuitive architecture. The front-end is built with react, tailwind and typescript, while the back-end is built with typescript and APIs like Hugging Face and OpenAI while being run on Vite.

Challenges we ran into

• Learning all the new APIs (Hugging Face and OpenAI), and dependencies in such a short amount of time to integrate seamlessly into our website. To overcome this challenge, we used existing AIs to prototype and refine our final product.
• Many issues related to state changes and audio detection issues. Audio detection issues led to either duplication or overwritten transcription. These were very challenging bugs but were solved with rigorous debugging.

Accomplishments that we're proud of

We are proud to create a functional website that helps users daily with learning. We always wanted to provide users with something that they can use as their companion for studying. Remained dedicated to this hackathon project and persisted through any obstacles we encountered.

What we learned

• The importance of collaboration to produce high-quality and complete code
• The efficiency and autonomy of AI models in repetitive or time-consuming tasks
• Learned to prioritize tasks in order of importance with strict deadline

What's next for Study Engine

• We can improve the accuracy and quality of audio transcription along with AI-generated answers
• Increase the summary size by fine-tuning the custom models
• Add transcription for different languages
• Implement a database system to store and retrieve the old transcription files and Q&A data

Built With

• html
• huggingface
• javascript
• jsx
• react
• tailwind
• typescript
• vite