Moderation engine

AI-assisted moderation, designed for medical discussion.

This platform is not ordinary social media. The moderation layer is built to filter spam, scams, herbal influencer rubbish, fake certainty, dangerous advice, defamatory posts, and emergency-delay content.

Three-layer moderation model

1. Hard block

Spam, fake cures, WhatsApp / Telegram selling, repeated promotions, DM harvesting, impersonation, and obvious junk never appear publicly.

2. AI scoring

Posts are scored for medical danger, misinformation, commercial intent, disruption, legal risk, and emergency signals.

3. Human review

Grey-zone cases, rare unusual presentations, legal risk, clinician disputes, and account penalties go to moderators.

Auto-hide examples

“Stop your anticoagulant.”
“Treat sepsis naturally.”
“Doctors don’t want you to know this secret cure.”
“DM me for the real treatment.”
Supplement / herbal / affiliate pushing.

Emergency escalation

Chest pain
Stroke symptoms
Collapse
Major bleeding
Severe breathlessness
Possible sepsis

Moderation scoring logic

Medical danger: 0–30 Misinformation: 0–20 Commercial / scam: 0–20 Disruption / behaviour: 0–10 Legal risk: 0–10 Emergency signal: 0–10 0–20 = publish 21–40 = publish with warning 41–60 = hold for review 61–80 = auto-hide + review 81–100 = block + flag account