J0RN@L

The J0RN@L doesn't just know you, it understands you

Comment

Inspiration

The insperation for this project is that people who suffer with mental health issues don't even know about it and end up not asking for help and let the pressure build. To the patient they cant tell that the Mental Health issues exist, but there are signs. Sleep hours, activity levels, and grades are indecators of poor mental health. Thus we want to use these factors to locat a deteriorating mental heath and help fix it before it grows. The way we want to locate them is using a log of a persons daily life.

What it does

This website takes daily reflection question answers and daily journal logs and uses them to understand ones mental health. It notices changes in pattern and overall unhealthy behavior. It uses the journal to get an overall understanding of ones day to day life aswell as note changes in how someone is writing. The reflection questions help pinpoint specific oddities it notices and helps the website get a better picture of you. The reflection questions are interactive with sleep data and other data that is related to the mental health so the website can find the strongest solution it.

How we built it

The UI was built and refined with google stitch, and then was later exported to VSCode. Then, we used Freebuff, a free alternative to CLAUDE CODE, to make it funcitonal and link the pages. We used supabase for authentications and db, and we used blackbox ai for the llm. We used codex for debugging and Nvidia NIM for getting a STT API key. Everything stayed free for us, within our budget of $0.

Challenges we ran into

Some challenges that we ran into was that ui was hard to get just right and we had trouble matching it to the vision that we had. We tackled this by manually designing a color pallet and using google stitch to help visualize the vision and perfect it until we began implementing into the website. Another challenge that we ran into was the voice to text feature that we added to make Journelling easier and not feel like a chore. We struggled getting the Nividia API to use voice to text because the base url was hard to find so the feature was not working. We used the assistance of codex to figure out why it wasnt working and it through guess in check made the feature viable by making the url easier to find.

Accomplishments that we're proud of

We are proud that we got down a solid pitch even though both members were very nervous, through a lot of trial and error we perfected our 2 min pitch. We are proud of the Mental reaserch we did to build the most helpfull bot when it comes to giving feedback on lifestyle to help those who need it

What we learned

One thing we learned is what an api is and how to impliment it in our code by using api to analyze the different journal entries. We also learned how to use voice to text which we implimented throughout our code for the all text boxes.

What's next for J0RN@L

We want to add more ways to journal instead of voice and text, one way which we were considering was physically writing in a notebook and taking a picture. This would allow for more inclusive venting so more people feel comfortable. Another change that would advance our product is working with lisence Therapists to help the ai examine data and perscribe the best solution to the problem. While the ai is smart and is using the the solutions that we gave using reaserch, ofcourse a lisense Therapist will be benifical and helpfull to the user.

Built With

Share this project:

Updates