CASE STUDY
TwelveLabs and Dyn Sport —
Revolutionizing Sports
Content Creation
Customer profile
Dyn Media is an innovative streaming platform focused on delivering content experiences to diverse fan communities across multiple, quickly growing sports. Their mission is to increase the reach and appreciation of those sports and together with their fan base carry the positive values of sport into society. Across multiple sports, genders and competitions Dyn’s ambition is to tell the story close to the heart of fans at scale. To do so Dyn needs to be able to not just automate highlight production but be able to find the moments in a match that go beyond sports data.
Executive summary
In the fast-paced world of sports media, engaging diverse fan bases with emotional storytelling has become a critical competitive advantage. This case study examines how the team at Dyn partnered with TwelveLabs to implement cutting-edge video AI that transformed their content creation and distribution workflow. By leveraging TwelveLabs’ video understanding models, Dyn can now rapidly identify, extract, and repurpose key moments from thousands of hours of sports footage, getting back valuable time and allowing creators to focus on their creating and storytelling.
Challenges
Dyn Media faced several challenges in their content operations:
As a startup company in the media stratosphere with limited resources available, Dyn needed to find a way to manage the vast amount of content created while covering over 3,000 live events per season. Reducing the dependency on manual search and human memory of what happened when was one of the key ideas behind partnering with TwelveLabs.
Content Volume
Managing thousands of hours of sports footage from multiple leagues, tournaments, and events across a multitude of sports.
Resource Constraints
No resources for manual tagging and clip selection creating bottlenecks in the content pipeline.
Cross-Sport Coverage
Maintaining consistent quality across various sports with different visual elements and terminology.
Solution
Dyn Media partnered with TwelveLabs to develop an integrated
video content processing system with three key components:
Intelligent Video Indexing with Marengo
The team at Dyn implemented TwelveLabs’ Marengo2.7 video embedding model to create a comprehensive searchable index of their video library. The system:
Processes and indexes new content in quickly as games and events conclude
Creates multimodal embeddings that understand the content of videos beyond tags
Enables powerful natural language search across the entire content library
Identifies specific moments, actions, emotions, and interactions within the video content
Automated Content Generation with Pegasus-1
For generating descriptions, summaries, and captions, Dyn leveraged TwelveLabs’ Pegasus video-language model, which creates accurate, contextually relevant descriptions of video clips.
Supporting a Large Media Ecosystem
Dyn uses TwelveLabs AI to enable their editorial team to find, access and utilize relevant content in a faster manner. As Dyn has built a product to aggregate and distribute content from and with clubs, leagues and media partners – the metadata provided will support not just Dyn’s editorial team but through the Content Desk product will help teams that are awarded access to the content to improve their workflows as well. By doing so Dyn helps everybody to tell the best story possible to their target groups.
Key use cases
The implementation enabled several transformative use cases:
As a startup company in the media stratosphere with limited resources available, Dyn needed to find a way to manage the vast amount of content created while covering over 3,000 live events per season. Reducing the dependency on manual search and human memory of what happened when was one of the key ideas behind partnering with TwelveLabs.
Archive Activation
Previously underutilized archival footage has become a valuable asset as the system can now connect historical moments with current events, creating rich narrative threads that enhance storytelling.
Emotional Storytelling
Using TwelveLabs’ capabilities, Dyn Media now have the opportunity to add qualitative data to historical and current content allowing editorial users to find and tell the story that touch the heart of fans in a way only sport can do.
Timely Highlight Creation
Content teams can now generate highlight packages within minutes of events occurring. By using natural language search queries like “best Kempa goals in the second half of the match” or “emotional fan reactions after game winning goal,” editors can rapidly assemble compelling content without manually reviewing hours of footage.
Conclusion
The partnership between TwelveLabs and Dyn Media demonstrates the transformative potential of advanced multimodal AI in the sports media landscape. By implementing TwelveLabs’ video understanding technology, Dyn Media has not only overcome critical resource constraints and challenges but has also laid the groundwork for future use cases when it comes to empowering editorial teams. As both technologies and strategies continue to evolve, this collaboration sets a new standard for how sports content can be created, distributed, and experienced in the digital age. This case study illustrates that the future of sports media lies not just in content volume but in the intelligent, timely, and personalized delivery of the right content to the right fans—a vision that TwelveLabs and Dyn Media are bringing to reality.
