Nora AI

Home page Hero
Marketing Page 1
Marketing Page 2
Marketing Page 3
Dashboard
Dashboard
Dashboard
Scheduling Email
Scheduling Page
Lesson Notes PDF

Inspiration

Some people believe AI could harm education while others believe it could help improve it. We believe AI will improve education by a large margin. Taking advantage of AI tools, we wanted to genuinely improve how people learn all over the world. Now we are at the center of that revolution with Nora.

What it does

Nora is a study companion that allows users to learn about any topic through discussions with AI tutors. The tutors have realistic looks, voices, and mannerisms so the user will feel like they are talking to a friend.

A unique feature of the AI tutors in Nora is that they will frequently ask questions to the user during discussions. This is to maximize user engagement with the material. This mode of learning has been proven to be very effective as it maximizes user retention of any material (backing research).

Additionally, with the app a user can do the following:

Enable subtitles during calls
Schedule learning sessions
Upload study materials the tutor will lecture on
Download study notes based on the discussions
Track subjects and time spent studying

How we built it

After the ideation stage, we built a wait list and shared it. We did this to determine if there was interest in the product. The email list we quickly got showed us clearly that there was interest in the product. We then used Bolt to start vibecoding the front-end of the project (using React). We used Tavus API for video generation, Assembly AI for generating conversation subtitles, Supabase for persisting data, and we hosted the front-end on Netlify. We used git and github to collaborate as a team.

We also had to build a backend server to handle the subtitle transcription, and some other heavy operations our app needed. This includes generating learning materials from attachments a user sends in the app, among other things. The Back-end server was also built with Bolt.

Challenges we ran into

The most challenging (but equally most rewarding) challenge was setting up a custom video call implementation with Tavus. Implementing transcripts were also a bit challenging. It was also tricky collaborating with each other on bolt but we eventually devised a method to achieve our goal.

Accomplishments that we're proud of

We are proud that we have a fully functional app at https://noratutor.xyz, that has the features we set out to accomplish when we started this hackathon. It was a challenging but equally rewarding experience for us.

What we learned

We learned that AI tools can go a really long way to improve education. We also learned how to really use bolt to build robust applications. We explored all of its functionalities and mastered the tool. Bolt is a tool we will recommend from now henceforth.