Identify harmful content and/or ethnographic patterns through our moderation army!
GET IN TOUCHThe web is dark and full of terrors!
The evolution of the internet birthed a near-constant threat of harmful content. Content moderation, in the popular sense, is the practice of monitoring, filtering, and assessing content based on rules in eCommerce, review sites, online communities or social platforms that incubates a large volume of user-generated content (UGC). Examples of such content includes (but not limited to) violence, hate speech, obscene content, profanity, sexual language, inappropriate imagery, substance abuse, nudity, racism, religious or political sentiments, extremism, conflicts, fake news, scams, mental health concerns, and PTSD.
Everyday, the world generates 500 million tweets, 294 billion emails, 4 million gigabytes of Facebook data, 65 billion WhatsApp messages and 720,000 hours of new content on YouTube. With the generation of data, comes the promise of analytics where our content moderators are commissioned to detect trends, insights, and patterns to support research, market scoping, and/or product development.
We provide several types of content human moderation services that take leverage of contextual understanding and intelligence - something that isn't widely available in automated tools. Our team works with clients to setup workflows and quality parameters unique to their needs.
Image moderation.
Human moderators evaluate user-generated images in social platforms, forums, or other online communities and tag by violation, quality, or risks.
Video/Audio moderation.
Assess and flag audio-visual content based on predetermined criteria - enabling real-time detection of harmful, offensive content and trolls.
Text
moderation.
Commission our moderators to assess, translate, and flag text in documents, forums and online discussion boards, chats, and comments/ reviews.
We have prepared our workforce to cover multiple languages based on context with coverage of Bangla, Bulgarian, Chinese, Czech, Danish, Dutch, English, Estonian, Finnish, French, German, Greek, Hungarian, Indonesian, Italian, Japanese, Korean, Latvian, Lithuanian, Norwegian, Polish, Portuguese, Romanian, Russian, Slovak, Slovenian, Spanish, Swedish, Turkish, and Ukrainian.
Tools we generally use.
We co-design and develop workflows unique to every client with provisions of training and procurement of tooling solutions. General steps are as followed
Engage with client build on the problem case, analyse context and develop a cohesive brief covering parameters of moderation components, responsiveness expectations, performance indicators and list out outliers.
We arrange for training and regimentation of moderators and also provision for any special requirements, such as recruitment of domain experts, as required by clientele. An example being provision of a specific online training to build contextual understanding.
Procure content moderation tool(s) and develop structures for reporting, quality assurance, programme management and escalation, and establish milestones of achievements.
Integrated feedback loop in setup to activate real-time monitoring and evaluation, spearhead case-by-case investigations of edge-case and error patterns, and continuously improve model of service delivery.
Deliver reports in agreed periods as per delivery requriements as well as packaged data. Our reporting entails both qualitative and quantitative insights, patterns identified, and general work process related information.
All of our workforce are certified in CDC/NCHS Confidentiality Training to PIIs. This is critical for medical data and is also interoperable to banking and retail use-cases.
Our operational leaders are certified in Total Data Quality from the University of Michigan - adopting best practices to assimilate into client's use-cases and inform on possible data and bias issues.
Our moderators have undertaken and passed Content Moderation courses offered by Microsoft, inclusive of intermediate modules of classification and moderation of text using Azure.
on-demand workforce scaling potential
Based on requirement and business value, we can scale our on-prem and remote workforce by upto 30% of its current capacity. Examples include increasing workforce by 13% in a week's notice.
Get expert workforces to attend to your content moderation or research needs - be it text, audio, video, or image formats.
GET IN TOUCH