Voice Reflections

Inspiration

Our voices are our primarily means of communication, what if we could intelligently leverage them to reinforce self-learning and information retention? Our team at VoiceRE was motivated by this idea and continued t the development of a self-learning tool that not only tracks reflections but augments

What it does

VoiceRE is a self-learning and information retention tool that allows users to engage with material that they've been taught at a deeper level. They are prompted to self-explain topics to themselves in a short audio file which they are always able to listen back to. Additionally, VoiceRE leverages generative AI to develop test questions based on the reflection that the user inputs. All in all, our tool is very scalable and allows for a one stop shop when studying for long sessions or reviewing.

How we built it

We built this by leveraging many of the latest cutting edge technologies and platforms. We use Next14 as full-stack with Typescript and Tailwind and a Node.js backend. A duo backend server in Flask was leveraged to allow the use of OpenAi API and Google API in Python scripts. Our audio files were recorded live and dynamically sends it to a Cloudinary media explorer storage which then sends the link of the audio file stored in there to MongoDB. Mongo was the main database that stored not only the link information but also user information gathered from the user's experience. To protect user data and provide exceptional personal user experience, Auth0 was leveraged to ensure such to the greatest degree.

Challenges we ran into

We ran into challenges posting VoiceRecordings to our Cloudinary media storage as well as our transcriptions to our MongoDB, in tandem they are used to prompt our LLMs to generate questions

Accomplishments that we're proud of

We're proud of the fact that we continued persevering despite not being able to curate an idea in the first day.

What we learned

We learned how to use Flask as a backend and how all in all, it is better to completely write the backend in one language rather than having multiple.

What's next for Voice Reflections

After this hackathon, the team plans on creating new iterations to our project to ensure that it functions as we envisioned, additionally we want to add more support and use cases with the data that the user provides.

Built With

• auth0
• cloudinary
• flask
• mongodb
• next.js
• node.js
• tailwindcss
• typescript