Vionlabs Plex

Plex Success Story

At Plex, we’re always looking for ways to improve the media experience for our user experience and engagements.
Vionlabs technology provides some truly unique capabilities that we believe will be invaluable in helping us achieve our goal of making streaming content even more accessible and enjoyable for everyone.

Shawn Eldridge, Vice President of Business Development and Content at Plex

https://www.plex.com/
Location: United States
About: Plex is revolutionizing the streaming landscape with its innovative ad-supported video on demand (AVOD) service and free-to-stream live TV channels. With a rich content repository stemming from collaborations with industry giants like Warner Bros. Television Studios, MGM, Lionsgate, Legendary, and many others, Plex provides its users with a diverse range of viewing options.

By the Numbers:

  • A thriving community of over 25 million global users
  • A vast catalog boasting over 50,000 on-demand titles
  • An impressive roster of more than 300 live TV channels

The Challenge

With a rapidly growing on-demand catalog boasting 50,000-60,000 titles, Plex was at the forefront of providing diverse content. However, a lurking issue was the inconsistent metadata quality that plagued their extensive library. As Plex continued to venture deeper, the discrepancies became more pronounced. The challenge wasn’t just about consistency; it was about unlocking the potential of hidden gems within the collection. A large part of the content library was basically unused due to inconsistent or inadequate metadata.

The Solution

Plex Vionlabs Collaboration: The integration of Vionlabs’ Fingerprint Plus emerged as the game-changer Plex was seeking.

Achieving Metadata Consistency: Vionlabs’ solution ensured that movies were categorized uniformly, taking into account genres, sub-genres, moods, and emotions. This systematic approach eradicated previous inconsistencies and provided a reliable foundation for content recommendations.

Tailored User Experience: Plex has always prioritized the viewer experience. With Vionlabs Mood Vector, users could now effortlessly align their current mood with content that resonates, ensuring more efficient and satisfying content selection.

Optimized Content Recommendations: Armed with Vionlabs consistent metadata, Plex’s machine learning algorithms became more proficient. They could sift through the vast content library with precision, leading to an enhanced user experience and boosted content discoverability.

Listen to what Scott Olechowski the co-founder and Chief Product Officer at Plex has to say about the collaboration with Vionlabs:

The integration of Vionlabs’ technology into Plex allows users to enjoy an enriched viewing experience through features such as:

  • Fingerprint embedding can be utilized by data scientists to innovate around personalization, content analysis, and more
  • 30+ Predicted Genres for Plex to categorize each title that can be combined with other nuances in the data such as mood and keywords to create many micro-taste groups
  • Measurement of the emotion and mood of each scene or character with unprecedented accuracy and can recognize 700+ different mood tags inside a single video file.
  • 20+ Mood categories extracted by AI, allowing Plex to connect titles across different genres.
  • Classification of each Mood Tag into a more general category so that content will be accurately categorized.
  • 1600+ descriptive keywords for the content story with weights reflecting the story of the content from different perspectives resulting in relevant keywords, describing the main topic or elements of the content. This allows Plex to create detailed personalization around topics, genres, and more, and significantly improves the search and discovery experience to keep the viewers engaged.

The Support

A Seamless Collaboration. Vionlabs’ unwavering support ensured that the integration process was smooth, scalable, and cost-effective.

 

Want to know more?

Your might also like…

No Results Found

The page you requested could not be found. Try refining your search, or use the navigation above to locate the post.