Your Safe Space

This chatbot provides a safe space for teens to talk about their struggles while detecting early signs of addiction. If risks appear, it privately notifies doctors so they can step in sooner.

Comment

Inspiration

  • We’ve seen friends struggle with drug addiction firsthand.
  • Their experiences made us realize that addiction often starts subtly with stress, loneliness, or social pressure long before it becomes visible.
  • If they had received help earlier, recovery could have been much easier.
  • That inspired us to build a system that can gently recognize those early signs.

What It Does

  • Our project is a mental health chatbot designed to help identify early indicators of drug addiction and emotional distress through natural conversation.
  • Teens can chat with the AI like they would with a friend
  • The AI listens, learns, and picks up on subtle cues about mental health and substance use
  • If the system detects warning signs, it sends a private alert to the doctor (not shown to the user)
  • Once the conversation ends it analysis the user and provideds a file with a detailed diagnosis.
  • The doctor can then review potential cases and provide timely care
  • The idea is to make it easier for patients, especially teens to express themselves openly, without fear of judgment.

How We Built It

  • Frontend: HTML, CSS, JavaScript (chat UI + interaction flow)
  • AI Backend: Python (Google Colab), using machine learning models trained on: Students Drugs Addiction Dataset 2024 (Kaggle) ##Logic:##
  • Detect keywords related to substances (e.g., “weed”, “ecstasy”, “fentanyl”)
  • Analyze mood and language cues
  • If triggered, provide helpful resources (CVC, SAMHSA, etc.)
  • Notify the doctor dashboard for review

Challenges

  • Managing merge conflicts during collaboration
  • Making the chatbot feel human, not robotic
  • Balancing data privacy and medical responsibility

Accomplishments

  • Built a chat system that actually feels supportive
  • Integrated mental wellness checks
  • Designed a resource response system for drug-related mentions
  • Created the foundation for an AI-powered early detection tool that could save lives

What We Learned

  • Addiction has many mental health indicators — social withdrawal, financial issues, sleep changes, etc.
  • Teens are more honest with nonjudgmental platforms
  • Early detection matters and empathy in design is just as important as the code itself

Whats next for Your Safe Space

  • Expand detection to other mental health issues (e.g. depression, anxiety, eating disorders)
  • Improve emotion recognition using sentiment analysis and contextual language models
  • Build two chat modes: one for teens (friendly tone), one for doctors (data-focused summary)
  • Partner with hospitals or school counselors to pilot real use cases
Share this project:

Updates