Prime Video has revealed details of generative AI solutions it is using to enhance the viewing experience for its Prime Video subscribers.
The streaming company has utilised Amazon Web Services’ (AWS) Amazon Bedrock to create a number of genAI tools. The streamer is using five different ways to deliver better experiences:
Personalised content recommendations: designed to ensure users spend more time viewing content than searching for something to watch, Prime Video’s AI-powered personalised content recommendations are placed directly within its ‘Movies’ and ‘TV shows’ landing pages. Users are presented with collections of content curated according to their interests and viewing history.
X-Ray recaps: trained using Amazon Sagemaker and powered by a blend of Amazon Bedrock managed foundation models and custom AI models, X-Ray Recaps builds on Prime Video’s existing X-Ray features by providing viewers with short summaries of seasons, episodes or sections of episodes. Detailed descriptions of key events, places, times and conversations are generated, with Amazon Bedrock Gaurdrails ensuring summaries are spoiler-free.
Prime Insights are AI-driven enhancements aiming to bring sports fans closer to the action. Built by Prime Sports producers, engineers, on-air analysts, AI and Computer Vision experts with an AWS team using Amazon Bedrock, the new insights illustrate hidden aspects of a game and forecast pivotal moments before they happen. For example, the “Burn Bar” on Prime’s NASCAR coverage combines live tracking data and telemetry signals to analyse fuel consumption for every car on track, identifying team fuel strategies and showing which drivers are conserving fuel during a race. Additionally, Rapid Recap provides a catch-up feature for fans joining during an event, automatically compiling a full recap of highlights which is dropped into their live feed.
Dialogue Boost analyses original content audio and identifies points where dialogue may be difficult to hear. The feature isolates speech patterns and enhances the audio to make the dialogue clearer. The AI-based approach improves accessibility by delivering a targeted approach to portions of spoken dialogue rather than providing general amplification.
Launched in English, the feature now supports an additional six languages including French, Italian, German, Spanish, Portuguese and Hindi.
Enhanced video understanding: Prime Video is using the Media2Cloud on AWS guidance, which delivers media analysis at frame, shot, scene and audio levels, enriching assets metadata and enabling teams to effectively track rights, verify QC, analyse and monetise their content. Powered by Amazon Bedrock, Amazon Nova, Amazon Rekognition an Amazon Transcribe, Media2Cloud enables enhanced content management, boosting search capabilities and improving audience engagement. Asset metadata is automatically fed to AWS partner Ikonic. Prime Video has been able to enrich hundreds of thousands of content assets, improving discoverability in its marketing archive.
