The most contextually aware AI to help you detect abuse in its many forms and take action
13%
of U.S. adults abandon
platforms after seeing
harassment online
1 in 4
people who are targeted
abandon their online
presence
4.5x
Increase in negative
brand perception
75%
of online abuse
goes unreported
Abuse takes many forms, is always changing and context matters.
Trained on the dark web, gray web, and open web, Sentropy stays ahead of online abuse and detects what others can't find, so your community can have the safest experience.
Monitor, review and prioritize detected abuse with easily customizable queues.
Context and scoring make it easy to know when and where to focus your attention.
Every community is different.
Define and take the right actions for your community through customized webhooks. Integrate any type of action, from content to user management right in context of the queue.
Designed to integrate with your platform’s content with a few lines of code, you get value immediately, while Sentropy’s core is always learning from the latest forms of abuse across the web, so you can trust you’ll be protected into the future.
Sentropy is the only solution deep enough to find the biggest threats to your community
This Demo includes offensive text content.
Select one of the classes above to see examples of Sentropy's detection capabilities in action.
Our Financial Scam model detects content that seeks to deceive individuals into providing funds or sensitive information.
1
2
3
4
5
6
7
8
9
10
11
12
13
curl "https://api.sentropy.io/v1" \
-H "authorization: Bearer $TOKEN" \
-d '{
"id": "3r00dmv",
"author": "AndrewP",
"segment": "meetup",
"text": "I’d like to help you
out .All you gotta do is sign up
and I’ll send you money to put into
your forex account. Me or my team
will give you a signal and you make
money"
}'
Seemingly ambiguous phrases are regularly weaponized across the web. Our technology is trained to distinguish between innocuous and harmful ones. This phrase may appear to most people - or keyword detection systems - to be something related to cuisine. However, this is actually an anti-Muslim reference
1
2
3
4
5
6
7
8
9
curl "https://api.sentropy.io/v1" \
-H "authorization: Bearer $TOKEN" \
-d '{
"id": "c73hd8k",
"author": "zootywoot",
"segment": "league-chat",
"text": "i want our country back.
remove kebab"
}'
Our Illicit Drugs model detects content that conveys an intent to purchase, sell, cultivate or manufacture illegal drugs, controlled substances, and/or unauthorized prescription drugs.
1
2
3
4
5
6
7
8
9
curl "https://api.sentropy.io/v1" \
-H "authorization: Bearer $TOKEN" \
-d '{
"id": "dnrlq5b",
"author": "anon9999",
"segment": "voice-chat",
"text": "testosterone propionate pow.
99% purity"
}'
Misspellings and character substitutions - whether intentional or unintentional - should not prevent detection systems from identifying abusive content. Our AI is trained to detect abuse even when it's hidden in plain sight.
1
2
3
4
5
6
7
8
curl "https://api.sentropy.io/v1" \
-H "authorization: Bearer $TOKEN" \
-d '{
"id": "jbhtyvl",
"author": "Princess_Howell",
"segment": "gathering-hall",
"text": "F--- you you @.$$"
}'
Every community is unique. This phrase would be unacceptable on a dating site but likely acceptable on a first-person shooter gaming platform. Sentropy's technology can be customized to match each community's needs.
1
2
3
4
5
6
7
8
9
curl "https://api.sentropy.io/v1" \
-H "authorization: Bearer $TOKEN" \
-d '{
"id": "jrcmsut",
"author": "fallax999",
"segment": "general",
"text": "i'm gonna put a bullet in
your head!"
}'
Misspellings and character substitutions - whether intentional or unintentional - should not prevent detection systems from identifying abusive content. Our AI is trained to detect abuse even when it's hidden in plain sight.
1
2
3
4
5
6
7
8
curl "https://api.sentropy.io/v1" \
-H "authorization: Bearer $TOKEN" \
-d '{
"id": "6dflo42",
"author": "SterlingArcher",
"segment": "general-chat",
"text": "He’s such a fkking idiot"
}'
Physical harm is not always a threat against others. Sentropy helps communities identify vulnerable users that are in need of help.
1
2
3
4
5
6
7
8
9
curl "https://api.sentropy.io/v1" \
-H "authorization: Bearer $TOKEN" \
-d '{
"id": "9u24bch",
"author": "TheAmbulanceMan",
"segment": "confessions",
"text": "it's too much i want to end
it all"
}'
Threats of rape don't always inclue the word "rape". These would be missed by many detection systems. Sentropy's technology recognizes the unconsentual nature of this message.
1
2
3
4
5
6
7
8
9
curl "https://api.sentropy.io/v1" \
-H "authorization: Bearer $TOKEN" \
-d '{
"id": "hrjjn4e",
"author": "PapaSmurf",
"segment": "selfies",
"text": "i'm excited to have sex with
you whether you want it or not"
}'
White supremacist groups employ a diverse numerical lexicon to communicate amongst themselves. Substituting the number 8 as a stand-in for the letter H (the eighth letter of the alphabet) allows the string 88 to become shorthand for Heil Hitler.
1
2
3
4
5
6
7
8
curl "https://api.sentropy.io/v1" \
-H "authorization: Bearer $TOKEN" \
-d '{
"id": "p33bjnv",
"author": "John_a_Patriot",
"segment": "free-speech",
"text": "here for the 88"
}'
detect abuse on your platform
Find abuse through contextually aware Al
Integrate easily with a few lines of code
Trained on the dark, gray and open web
Deepest abuse detection available
defend your users
Review content with rich context
Take action through custimizable webhooks
Tag, get notified, and manage content with ease
Protect your communitY
Expert, purpose-built tooling
Adapted to your community
An evolving solution for tomorrow’s challenges
Focus on user safety, not tech maintenance
