Socha

Journaling for the Social Age: Bridging the gap between individual reflection and communal support

Comment

Inspiration

Mental health is often treated as a private, solo journey—yet human connection is one of the most powerful tools for emotional resilience. I was inspired to build a platform where the privacy of a personal journal meets the strength of a social circle, ensuring that no one has to process their day alone.

What it does

Socha is a social-emotional tracking platform that encourages daily reflection.

  • Intelligent Journaling: Users record thoughts and moods, which are analyzed in real-time.
  • Sentiment Analysis: Each entry is processed to determine an emotional score $S$, where $S \in [-1.0, 1.0]$. A score nearing $1.0$ indicates positive sentiment, while $-1.0$ reflects distress.
  • Circles: Users organize into "Circles" (communities) where aggregated, privacy-first insights allow members to see the general emotional "pulse" of their group without violating individual privacy.
  • Acts of Kindness: The app suggests actionable ways to support friends based on the community’s collective state.

How I built it

I architected Socha using a modern full-stack approach:

  • Frontend: A cross-platform mobile app built with React Native and Expo. I utilized Apollo Client for state management and Three.js (via expo-three) for immersive data visualizations.
  • Backend: A Node.js and Express server serving a GraphQL API.
  • Database: MySQL managed through the Sequelize ORM to handle complex relationships between users, entries, and circles.
  • Intelligence: Integrated the Google Cloud Natural Language API to perform high-precision sentiment and magnitude analysis on user text.

Challenges I ran into

One of the primary hurdles was designing a GraphQL schema that could efficiently aggregate emotional data across different "Circles" without compromising performance or privacy. I also spent significant time fine-tuning the sentiment analysis magnitude to ensure that nuanced entries weren't "misread" by the AI.

Accomplishments that I'm proud of

I'm incredibly proud that Socha won First Place at HackSC 2020. Seeing the system successfully translate individual text entries into a cohesive community "health" dashboard was a major milestone for our team.

What I learned

This project was a deep dive into the intersection of NLP (Natural Language Processing) and social psychology. I learned how to bridge disparate tech stacks—connecting a mobile frontend to a cloud-based AI engine—while maintaining a focus on user-centric design and ethical data handling.

What's next for it

I plan to expand the "Circles" feature with more granular community goals and deeper integration of 3D emotional landscapes using Three.js to help users visualize their mental growth over time.

Built With

Share this project:

Updates