Online Safe Zone

Executive Summary

The AdvitCare Online Safe Zone is a curated, secure digital ecosystem hosted on Discord. Unlike public social spaces, this "Safe Zone" is a closed-loop community designed to provide academic support (tutoring) and peer connection in a heavily moderated environment. The primary goal is to foster a "Third Place" (neither home nor school) where students can study and socialize safely under the supervision of teachers and guardians.

Safety, Legal & Compliance

• The "Walled Garden" Policy (Entry Verification):

  • Step 1: Application: Students apply via the AdvitCare Academy web portal using a valid School Email (.org) or Parent/Guardian ID.

  • Step 2: Identity Proofing:

    • Students: Confirm their DOB, current school and state.

    • Parents/Guardians: Submit a letter of approval.

    • Tutors: Mandatory background checks (Criminal Record/Sex Offender Registry) before receiving invite codes.

  • Step 3: Unique Tokens: Each user receives a one-time-use invite token linked to their real identity in the backend database.

• COPPA & Age Compliance:

  • Discord Limitation: Per Discord Terms of Service, users must be 13+.

    • Grades 7-12: Full access to the Discord Community.

    • Grades K-6: cannot legally be on Discord. They are routed to a separate "AdvitCare Junior" portal (web-based whiteboard/chat) accessed via parent login.

  • Data Minimization: No recording of voice channels unless explicitly triggered for safety reviews. Chat logs are retained for 30 days for safety auditing, then anonymized.

• Zero-Tolerance Code of Conduct:

  • No Bullying/Hate Speech: Immediate ban and contact to parent/Guardian.

  • No DMs (Direct Messages): The server is configured to disable DMs between members to prevent grooming or unmonitored harassment. All communication happens in public channels.

  • No adding: The server is configured to disable the ability to add each other as friends.

  • Academic Integrity: No cheating, answer-trading, or plagiarism.

• Operational Roles & Structure:

  • Student (Verified): Access to study lounges and ticket creation.

  • Tutor (Verified Pro): Background-checked adults or high-achieving university students. Identifying color: Gold.

  • Sentinel (Moderator): 24/7 paid staff members responsible for safety, not tutoring. Identifying color: Blue.

  • Watchdog (Bot): Automated AI moderation tools.

• Physical Server Layout (Channel Architecture):

  • #arrival-gate: The only visible channel for new users. A bot verifies their token before unlocking the rest of the server.

  • Subject Wings (Text ): #math-help / #science-lab / #humanities-hall.... etc.

  • #chill-lounge (heavily filtered casual chat).

  • #pet-pics / #art-gallery.

  • The Help Desk (Ticketing): Students open a "Ticket" to request a 1-on-1 tutor session. This creates a temporary private channel visible only to the Student, the Tutor, and the Sentinel.

• Tutoring Workflow (The "Help" Function):

  • Request: Student uses a slash command (e.g., /tutor math "Help with Calculus") in the #help-desk channel.

  • Assignment: The "TutorBot" pings available math tutors. A tutor claims the ticket.

  • Session: A private text channel is created.

    • Safety Rule: A "Sentinel" bot sits in the channel recording the text for safety assurance.

  • Closure: Once the session ends, the channel is archived and the transcript is stored for 7 days in case of disputes, then deleted.

• 24/7 Monitoring Strategy:

  • Safety is maintained through a hybrid of AI and Human oversight.

• AI Automod (First Line of Defense):

  • Profanity Filter: Hard-blocking of slurs, curse words, and sexually explicit language.

  • Image Scanning: All images uploaded are pre-scanned by AI (e.g., Cloud Vision API) for inappropriate content (NSFW/Gore) before being posted.

  • Anti-Spam: Prevents link flooding or scams.

• Human Sentinels (Second Line of Defense):

  • Shift Rotation: Staff operate in 8-hour shifts to ensure a moderator is always online, regardless of timezone.

  • Grooming Detection: Sentinels are trained to spot "grooming" behaviors (e.g., an adult tutor asking a student for personal photos or to move to Snapchat/Instagram). If detected, the tutor's information is reported to local authorities, they are removed from the community and banned.

• Crisis Escalation (Mental Health Integration):

  • The "SOS" Command: If a student types /help or keywords like "suicide," "hurt," or "depressed":

    • The message is auto-flagged for Sentinels.

    • The bot immediately DMs the student a "Crisis Resource Card" (Hotlines, Text Lines).

    • No Counseling Rule: Tutors are strictly forbidden from acting as therapists. They must immediately tag a Sentinel who will escalate to the professional administration team or local authorities if necessary.

At AdvitCare Academy, we believe kindness is key to creating a safe learning environment. Join other students and families across the world in spreading kindness every day.