Flow

The harsh reality is that 26% of teenagers in Canada struggle with their mental health. Flow set's out to create a safe, calming environment for teens who are struggling with their mental well being.

Comment

Inspiration

As the youngest contestants in the hackathon, we wanted to make something that we could see ourselves using. Not only are we Canadian, but we are also teenagers. Unfortunately, 26% of teenagers struggle with the mental health, and rate their own mental health as poor. We set out to make a safe space for teens to be calm, heard, and understand that they matter. Flow is meant to be a space for teens to release, and "Flow" into finding the calm within the storm, or the peace within the rough times.

What it does

Flow is essentially a central page to calm you down, and make you feel good. Flow has 3 main abilities. The first is sound, then quotes, and finally chat.

How we built it

The sound section is a set of sliders associated with different frequencies. Brown noise, white noise, and pink noise. Brown noise is a deep frequency, that sounds like a low rumble. White noise is medium pitched frequency, and finally pink noise, a mix of all the frequencies audible to the human ear. The sliders affect the volume of the different noises, and allow you to customize the combinations. Not only this, but you can also create your own ambient noises using the text to sound effect tool from eleven labs.

Next is the quotes. For the quotes, the gemini api is prompted to generate 10 motivational quotes that carry meaning. Then, these quotes are turned to audio with eleven labs text to speech.

Finally, is the chat. We did some research, and spoke to other AI's about mental health and self esteem, and they all treated us like patients. We have prompted the gemini api to treat you like a friend, and lay out boundaries, as to not pry for too much information. This chat can also become a voice chat using eleven lab's text to speech for the AI end, and speech to text for your end.

Challenges we ran into

We ran into challenges setting up the voice chat. At first our API key's kept on failing, and getting flagged, and then we found the voice for the AI was far too robotic to feel soothing and comforting. We spent a lot of time learning how to change the voice and ended up making it a choice for the user.

Accomplishments that we're proud of

We are really proud of what we have made, considering we are the youngest members of the hackathon. We are most proud of our main feature, the voice chat with the AI. This was really hard to get working, especially setting up the different voices, and the speech to text.

What we learned

We learned to keep trying. Imagine you are a swimmer, swimming against the tide. This is hard to do, but you still push through. Many of the issues we ran into were with the API's and not the code itself, and we eventually kept trying and trying until it worked.

What's next for Flow

We plan on adding an account system, so you can save the audio you generate, and so the ai chat's can get more personalized over time.

Share this project:

Updates