Using AI to automate content moderation and compliance with
summary
Shows a concrete AWS moderation stack that actually runs: S3 staging, Step Functions orchestration, Rekognition for DetectModerationLabels and segmentation, Transcribe for audio, Comprehend and Translate for language, Textract for documents, A2I and Ground Truth Plus for humans and labeling, plus EFS mounting to avoid repeated video downloads. Claims include reclaiming up to 95% of manual moderation time, Flipboard processing ~90 million images/day and catching ~63,000 violations/day, Media2Cloud demo deploys in ~20–25 minutes and image checks return in about one second.
AI-written summary · about
chapters (15)
- 0:00Introduction and agenda
- 1:06User-generated content era impact
- 9:08AWS moderation reference architecture
- 17:49Flipboard content moderation case
- 19:18Video moderation workflow demo
- 22:09MobiSocial gaming moderation case
- 23:58Audio moderation strategies
- 27:15DeNA live audio moderation
- 29:05Multilingual text moderation
- 30:10Rich text document processing
- 33:30AI services for content moderation
- 35:22AWS Content Analysis solution
- 40:03Media2Cloud demo walkthrough
- 47:38Q&A on moderation and recognition
- 54:29Getting started with AWS AI services
related talks
16mFighting Harmful Content with AI: Real-World Safety
12mUsing AI/ML to scale governance, risk management, and audits
55mAccelerating auditing and compliance for generative AI on AWS
01h 00mThe Future of Content: Context-Aware, Centralized, and Written by Robots
14mAI That Moves Business. Cloud That Scales
45mPanel: From Spam to Speech: Policing the Next Generation of "Unwanted Traffic"
21mIntelligent Workflow Automation with Lambda and AI Models
20mExplore what’s possible with new AWS generative AI services