Inspiration
Many people rely on ChatGPT for advice, questions, and a lot of personal things. I am also a part of this group of people. I thought it would be quite helpful to see how we really are as people through the lens of the tools we use the most when it comes to personal inquiries.
What it does
It uses the Cognitive Behavioral Therapy model to analyze your thoughts, emotions, and behavior. This is widely used by therapists all around the world. PsycheFlow gives you a report based on single entries and all of them combined. It tracks your mood over time and gives you an overall report of yourself.
How we built it
We built it using Tailwind and React. We used Claude's Sonnet 4.5 for the LLM integration. It has two separate API calls, one to give users a psychological report of themselves based on one entry, and another based on all of the entries combined.
Challenges we ran into
It was hard to figure out how exactly we could ask users questions that would be the most helpful for us to extract the needed information out of. It was also a challenge to make sure the prompt was tailored to be as accurate as possible.
Accomplishments that we're proud of
If you are fairly honest then the system does give you an accurate report of yourself. The system generates structured and consistent psychological reports that align with the principles of Cognitive Behavioral Therapy.
What we learned
It is really helpful to understand the true power of the tools you use daily.
What's next for PsycheFlow
Try deploying it for personal use as its applications seem very impactful in self-improvement. On top of this, also add user accounts to the data is stored in accounts and not just locally. Perhaps also experiment with different therapeutic frameworks for the psychological reports.
Log in or sign up for Devpost to join the conversation.