Liam Morrison

Want to increase the value of your archive, and supercharge your content discoverability and monetization? This chalk talk covers how to generate and make searchable, detailed metadata for large content catalogs (such as images, video and audio files, and documents) using computer vision, speech to text, and natural language processing.

It can be challenging to implement an ML-based approach to personalization in the media and entertainment industry, but those companies that do see significant benefits, including increases in video consumption and subscriptions. In this session, learn how Amazon Personalize can help you add ML-powered recommendation engines to your website using pretrained ML models built on years of ML...

Personalizing the user experience is proven to increase discoverability, user engagement and satisfaction, and revenue; however, many AWS customers find personalization hard to get right. Effective recommender systems require solving multiple hard problems, including constantly changing user behavior, new catalog items (cold start), etc.

While model creation continues to become more automated, creating good datasets is still a problem. No matter how advanced your model is, it will not be perfectly accurate all of the time.