NAB 2019 Reflections: Levels Beyond CEO Art Raymond on the Evolution of Reach Engine, AI, Esports Integration with AWS is key in the advance of the platform By Jason Dachman, Chief Editor Wednesday, April 17, 2019 - 1:45 pm
Print This Story
Story Highlights
At the Amazon Web Services booth at NAB 2019, Levels Beyond showcased the latest integrations of its Reach Engine Media Production as a Service (MPaaS) platform with AWS. Its partner portal at the AWS booth highlighted how Reach Engine 2.6 has integrated with Amazon Cognito, AWS Elemental MediaConvert, and Amazon S3 and S3 Glacier to meet the needs of modern media productions. Using these AWS tools, Reach Engine users can manage everything from media acquisition and project assignments to editing, versioning, formatting, distribution, cloud storage and more.
SVG sat down with Levels Beyond CEO Art Raymond to discuss the AWS integrations and latest advances in the Reach Engine Platform, the company's groundbreaking work with Fox Sports during the 2018 FIFA World Cup, how AI is streamlining sports-production operations, and why he believes Blizzard Entertainment and other esports organizations are reinventing the sports-media landscape.
Levels Beyond's Art Raymond: Our goal is to be like the systems that played a role in the regeneration of manufacturing plants in the 1980s, where they became much more efficient.
How is the Reach Engine platform evolving/improving to better serve M&E and sports-media industries?
Reach Engine has gone even further to become a subplatform toolset. Our goal is to be like the systems that played a role in the regeneration of manufacturing plants in the 1980s, where they became much more efficient. Those plants knew they needed planning tools for inventory, supply chain, assembly, and so on, so that everything would be laid out in advance. Then they needed to know the state of each of those stages and be able to track them reliably. By doing that, they were able to have greater speed and greater quality, as well as cut the fat and reduce the cost of manufacturing on a massive scale. That kind of thinking just hasn't been part of our [media] industry thus far.
Our focus has been to be that [system] and create those planning toolsets, which we call subplatform-management tools. In sports, you may have 300 games in a dozen languages [being distributed] on nine major platforms, which means there are literally thousands of jobs to do. Our job is to manage and help orchestrate that entire production cycle. [Reach Engine] is built to help you organize it, plan it, manufacture it, and assemble it - while automating as much of the background as possible.
What Reach Engine features and integrations are you highlighting at the AWS booth?
We have a full API integration with Amazon Cognito, which we think is going to really simplify things [for our customers], going between AWS and their own infrastructure.
Everything in [AWS Elemental] MediaConvert is now fully API-enabled as a connector, so it's totally native. In the middle of anything, you could set a workflow job up, pick a MediaConvert service from a dropdown menu, drop in it with a status monitor, and know exactly how it's progressing.
And all of the Amazon S3 and Glacier storage layers are now completely [integrated] with Reach Engine. You can just [operate] inherently across S3, Secondary S3, Glacier, and Secondary Glacier, and all the ID-tracking and asset-management data comes back to us natively. So we're truly integrated end-to-end with Amazon now, which is really great.
Can you dive a little deeper into Reach Engine's role for Fox Sports during the 2018 FIFA World Cup production?
The project definitely wasn't easy, but we learned a lot. We worked closely with EVS, Telestream, Aspera for transcoding and optimizing the transport of all those live streams. Then we took over the queuing in order to maximize the Aspera service, and it went into AWS in New York. At its peak, [the production had] 66 infrared camera feeds and 120 TB at peak-load flow. All of the visualizations of every single asset are in Reach Engine. And then every single remote editor - all 348 of them - could sit inside Adobe Premiere and see AI clips coming in, watch the streams coming in, switch between multiple streams, and edit on the fly. And they had the ability to [view the content libraries] inside a truck in St. Petersburg or [the IBC in] Moscow or [Fox Sports HQ] in L.A.
How are you looking to take the lessons learned from this project and deliver new capabilities to other organizations?
We learned so much on that project that we're now actually going back to other organizations to see how we can [leverage those workflows]. After that event, we went to Blizzard and some of the other gaming companies that we work with - who are really global organizations - and show them how we could lay out the infrastructure and reduce their costs significantly. We can control the operation using all remote resources and can control the internationalization in-country, and they thought that was incredible. For us, even though it was an extremely complex and difficult [project], all of these [workflows] are actually highly reusable, especially for sports.
How is Reach Engine leveraging AI to automate workflows and boost productivity for its users?
If you look at the AI tools that AWS has, every one of them is an amazing product, but it's like a sledgehammer and has a lot of little attributes. If you want to see how that works in the real world, we've got a great vision of that and live real-world implementations. For ingest of live feeds, the NFL, Fox Sports, Blizzard [Entertainment], and others are using AI elements and metadata to identify clips and then lay all that into a timeline. We have one of the few multi-object timeline tools
