NORMAL

Normal helps colleges identify rising student stress early through anonymous signals, grounding tools, and real-time dashboards.

Comment

NORMAL

Inspiration

In most campuses, student stress becomes visible only after something goes wrong.
Existing solutions focus on individuals, require identity disclosure, or depend on manual reporting—leaving institutions blind to early warning signs. We were inspired to build a system that makes silent pressure visible before it escalates, without violating privacy or turning mental health into a clinical problem.

What it does

NORMAL is an anonymous, AI-assisted campus awareness platform.

Students can:

  • Express stress anonymously through a guided chat
  • Use optional grounding interactions and reflective tests
  • Participate in a moderated peer forum

Institutions receive:

  • Aggregated, non-identifying stress signals
  • Time-based trend graphs and spike detection
  • Group-level insights without tracking individuals

NORMAL does not diagnose or treat—it helps institutions notice patterns early and respond responsibly.

How we built it

NORMAL is a full-stack web application built with a focus on privacy, clarity, and deployability.

  • Frontend: React (Vite) with a calm, accessibility-focused UI
  • Backend: Node.js with REST APIs
  • AI Layer: External AI API with strict prompt constraints to ensure non-diagnostic, reflective responses
  • Data Handling: Aggregated, time-based data only (no identity storage)
  • Deployment: Cloud-hosted web application with a working demo

The architecture is modular, allowing easy extension while maintaining ethical boundaries.

Challenges we ran into

  • Designing an AI interaction that is supportive but not advisory
  • Balancing anonymity with meaningful institutional insights
  • Avoiding feature creep while delivering a complete product
  • Maintaining ethical clarity under tight time constraints

Accomplishments that we're proud of

  • Delivering a fully working, deployed platform within the hackathon timeline
  • Building a deliberately constrained AI interaction for safety
  • Creating grounding tools that reduce cognitive load without gamification
  • Maintaining strong privacy principles through data minimization

What we learned

  • Simplicity and restraint often build more trust than complex features
  • AI systems in sensitive contexts must be intentionally limited
  • Aggregated insights can be more powerful—and safer—than individual tracking
  • Clear framing is as important as technical execution

What's next for NORMAL

  • Pilot testing with educational institutions
  • Enhanced trend analysis and alert customization
  • Improved moderation tools for the forum
  • Accessibility and multilingual support
  • Ethical review and compliance preparation for real-world deployment

Built With

Share this project:

Updates