Today, internet security company Sentropy unveiled its platform to detect and fight online abuse and malicious content. After years of witnessing hate speech and harassment drive churn and erode trust in online communities, the machine learning experts at Sentropy are helping online communities of any size protect their users. The company is introducing two new products, Sentropy Detect and Sentropy Defend, that provide easy access to proactive, bias-aware abuse detection through continuously evolving machine learning models.
“We built Sentropy because we all deserve a better internet,” said John Redgrave, co-founder and CEO of Sentropy. “Our children should inherit an internet that’s better than the one we have today. Until every platform has the ability to protect its users, we continue to risk the safety and security of our loved ones.”
The online communities people visit every day establish norms quickly, shift frequently, and can degrade rapidly. Millions of users are abused and harassed every day, but 75% of online harassment goes unreported. The inefficiencies of existing tooling have led companies to rely on human moderators around the world to manually review content. But even the most efficient human moderators – capable of reviewing 3,000 pieces of content each day at peak efficiency – are unable to keep up with the volume of abuse online. Despite widespread data demonstrating the mental and psychological trauma this job can inflict, the need to protect users has spawned a multi-billion dollar moderation industry that continues to be insufficiently equipped to address the problem and demands a more scalable solution.
"I’ve seen first-hand the difficulty of manually moderating online communities,” said Alexis Ohanian, co-founder of Initialized Capital. “The breadth and depth of this issue require serious resources and machine learning chops. Sentropy has built the tech that is much needed across social media communities. User safety has become a competitive differentiator for those willing to take a stand against abuse."
Sentropy Detect, the company’s flagship product which provides API access to its classification technologies, has been in private beta testing since June of last year. Today, Sentropy is making Detect available in a public beta along with the release of Sentropy Defend, a browser-based interface that enables professional-grade, end-to-end moderation workflows. Defend provides moderators with the intelligence and tools needed to identify specific forms of abuse, discover behavioral trends, and make data-driven moderation decisions that are more efficient and consistent. Sentropy’s products help to seamlessly monitor and protect digital spaces, powering discovery and communication that reduces abuse, drives deeper user engagement, and ultimately reduces churn.
“Abuse and harassment have been problems for online communities since the beginning of the internet,” said Megan Guy, partner at King River Capital, the lead investor in the company’s latest funding round. “What’s different now is the scale of digital communities and the increasingly important role they play in our lives. We believe user safety will become the security issue of our time, and we invested in Sentropy because this team and their technology are uniquely able to help companies protect their users.“
Sentropy built a machine learning platform embedded with deep bias research in order to provide actionable recommendations for addressing abuse. The platform leverages constantly evolving proprietary software that monitors the open, grey, and dark web for new behaviors and adapts to individual community guidelines and norms. Highlights of the Sentropy platform include:
- Quick-to-deploy API
- Dynamic, constantly evolving detection models
- Embedded bias mitigation
- Intuitive workflow tooling
- Actionable moderation recommendations
- Automated tuning to community norms
With Sentropy, users get the internet they deserve.
Source: https://www.sentropy.com/