Voice-first toxicity detection

Manage communication threats in real time.

Learn More

Shush is an AI-based stack that multi-tags communications using ML models.

The results can be used to detect threats and manage communities.


              Communication Data (voice / text) >
              Shush API >
              Labels JSON

Gaming and the metaverse
Streaming and mass communications
Call Centers

Abuse online

72%
of users have witnessed toxic behavior.
- Source: Harris Poll 2021

Online communications have all types of misbehavior that range from the offending to the threatening. Users don’t always know what to do, so they leave.

By integrating all communications into shush we can detect and auto manage cases helping you to protect your customer LTV.

A tiered solution

Leveraging on modern technologies to moderate using AI.

Analyzes voice sound waves and speech recognition to identify and flag aggressive/abusive behavior in real-time.

Machine learning applied to actions so the automatic rules adapt to moderator’s actions.

API

Audio and NLU processing on an easy to integrate API for multiple clients.


                  curl -X 'POST' \ $token \ 

                  -H 'accept: application/json' \ 

                  -H 'Content-Type: multipart/form-data' \ 

                  -F 'file=@audio.wav;type=audio/wav'

Multi-labels

For threat detection, user insights and community health.

Recording: 542435

Content: racism toxic words
User: male 20-30s angry
Others: doxxing phone

Dashboard

A web app to help your CS and CM teams deal with threat cases.

User: 2lkcxz2x

Threat level

.92

High

Doxxing

55%
Hate Speech

89%
Sexism

23%
...

A group of people of various ages happy with their phones

For a safer space

That everyone can enjoy

Improving online gaming scene by creating a safe space for others to share.

In console, PC and mobile gaming, taking care of sensibilities following current trends.