I’ve Seen The Future of The Streaming Video Tech Stack

I’ve said this before but it bears repeating: streaming is an evolutionary step in the television experience. Some of that has to do with how viewers can watch TV content now: from anywhere at any time, across different devices and screen sizes. The same content that was once shackled to the EPG by the cable operator is now free to be mashed up, aggregated, and organized in different and exciting ways.

But what that evolution really deals with is the technology underlying that experience.

In the world of broadcast television, it’s always been about hardware (although that is changing with many pieces of technology moving into the cloud). Conversely, in the world of streaming, it’s all about software. Any hardware components needed for the streaming workflow, such as encoders, have long been virtualized. The entire streaming video stack exists as software which can be broken apart and reconstituted in any process flow to suit the needs of the streaming platform or operator (it’s not that easy just yet, but I’m getting to that).

Of course, streaming is all about IP. And as broadcast continues to embrace IP over SDI, streaming and broadcast will continue to head on a collision course in the future where each will be based on the same technology and will just be seen as different ways to distribute content.

What that means is that the process for creating and ingesting content all the way to playback will be an entirely software-based workflow. And when the streaming tech stack software completely aligns with modern software practices, such as microservices and serverless functions, operators will truly have a modular, flexible, and entirely customizable way to build and deliver their service to users. Personalization won’t stop at the content but can, in this model, continue through to the very user experience itself: the entire interface and way in which users interact with the streaming service can be customized on the fly.

The trick in this new world, though, is enabling that kind of modularity and flexibility. Despite agile software development methods and improvements in development languages (such as Rust), it still requires a lot of software engineers to cobble things together. That’s where solutions like Norsk, from id3as—launched at Streaming Media East in May—will play such an important role. Sitting for a demo at NAB 2023, I quickly realized that I was seeing the future of how streaming services would be built.

That future is a simplified abstraction layer. Sure, abstractions have existed for a while in the world of SaaS. Just look at services like Segment. They provide a visual tool which takes the place of writing code to develop the integrations between different components within a specific process flow (in the case of Segment, a data pipeline). Ultimately, these visual representations simply handle what would otherwise be a manual coding process for connecting things together via API (as well as automatically update connections when APIs change).

But until now, nothing has existed like that for streaming. That’s because abstracting the video tech stack isn’t as easy as connecting sources and destinations when moving data around: components within broadcast and streaming often support APIs differently, there are elements of interoperability which require more than just programmatic connection, and the services themselves need to be provided.

That’s where Norsk is so intriguing. Although the abstraction layer is code-based (engineers could use a variety of languages), the code is simply about instantiation: basically, I want this player to pull in this video and provide this interactivity. What might have been thousands of lines of code were reduced to just a few. It’s not even hard to imagine another layer of abstraction like ChatGPT. How much more flexible would a streaming service be if an engineer could simply say, “I’d like an HTML5 player to pull in the source from URL XYZ with a bitrate ladder of ABC based on the user’s history”?

The evolution of the evolution of TV is coming soon, hinted at by services like Norsk. I for one can’t wait.

Personalization Is More Than What You Think It Can Be

With advances with generative AI, just-intime transcoding, SSAI stitching, and other streaming video tech stack components, companies like Infuse Video are demonstrating that the true vision of video personalization—changing the video content itself—is finally at hand.

04 Oct 2023

What the Growth of FAST Really Tells Us About Viewers

The popularity and growth of FAST show us that viewing behavior, despite the rise of streaming, hasn't really changed much at all. People want choice, but they want it in a way that meets their needs. FAST doesn't scrap the broadcast experience with which so many are familiar; it is evolving it in a way that broadcast could never do to improve upon the viewer experience.

09 Feb 2023

What's the Point of Streaming?

You might think it's crazy of me, as the executive director of the world's largest streaming technical association, to ask the question in this column's title. But I think it's actually important that we do a gut check every once in a while.

14 Nov 2022

Does Streaming's Future Success Depend On Discoverability?

The future of streaming is about unification built on top of a single open standard of content metadata. Right now, it's a matter of seeing the forest for the trees.

04 Oct 2022