Amazon Web Services (AWS) has launched AWS Elemental Inference, an AI agent designed to automatically transform live video footage for vertical screens in real time. This new service aims to help broadcasters and livestreamers better serve audiences on platforms like Instagram Reels, TikTok, and YouTube Shorts.
Previously, adapting traditional landscape video for vertical formats required extensive manual editing, often leading to delays and missed moments for viewers. AWS Elemental Inference analyzes video feeds and applies necessary transformations automatically, ensuring key action remains visible and maintaining a 9:16 aspect ratio. The service boasts a low latency of just 6-10 seconds, enabling near real-time optimization for vertical viewing.
Launching in select AWS Regions, the service integrates with AWS Elemental MediaLive, allowing users to leverage AI optimizations without altering existing video architecture. It is powered by managed foundation models, requiring no AI expertise from content creators.
Beyond vertical formatting, AWS Elemental Inference performs advanced metadata analysis, capable of detecting and extracting key moments like touchdowns or game-changing plays in live sports for immediate vertical highlight distribution. This technology allows for instant creation of optimized clips ready for social media.