Serendeepia

Amaritas

Keeping conversations safe and healthy

Amaritas protects the most vulnerable groups keeping online discussions healthy and productive for everyone

Notify me when the beta version is open

Detect harming speech

Language is a gift to humanity, but sometimes it can become a dangerous venom. Amaritas is an Artificial Intelligence service that automatically detects toxic speech:

  • Threats, Violence, Hostility, Bullying.
  • Insults, Resentment, Antipathy, Sourness.
  • Obscenity, Sexual Degradation, Nastiness.
  • Hatred, Rancour, Identity Hate.
  • Male Chauvinism.
Supported Languages

Amaritas is currently available for Spanish and English.

If you need help fighting harming or toxic speech in another language, please let us know.

// request
{
  "language": "spanish",
  "detect" : [ "TOXICITY", "INSULT", "THREAT" ]
  "text":

    "Yo no tengo la culpa de que seas una ignorante"
}

// response
{
  "score": {
      "TOXICITY": 0.98,
      "INSULT": 0.99,
      "THREAT": 0.01,
      }
}
Simple and easy to use API

Start using Amaritas within minutes thanks to our simple json API. You only have to create an API key and start sending requests. You can also start checking the integration with our mockup API.

Real Time Products

Amaritas allows you to keep safe and healthy online conversations at scale, supporting up to millions of concurrent conversations in real time. Relevant use cases include:

  • Online and workplace harassment detection and prevention.
  • Fighting against bullying in social media, schools, forums and other platforms.
  • Automatic moderation of forum discussions.
  • Early detection of sexting, grooming, violence and abuse.
  • Fighting terrorist indoctrination and hate speech.
  • Measuring the level of male chauvinism in conversations.