Welcome to Zixi Podcast. Today, we’re joined by Peter Hajittofi, CEO of Coralbay, and Gavin Smith, Director of Video Tech. Welcome to the podcast guys. Thanks for having us. Yeah. We’re gonna discuss how cloud playout is transforming broadcast workflows. CoralBay launched in 2016 with a vision to build next generation playout solutions that could run on premise or in the cloud. To kick off, could you share why CoralBay set out to create a cloud playout solution? Yes. In 2016, when we started, a lot of customers were using channel in a box solutions or, hardware systems. And what we found is those systems are are not suitable for the cloud. They don’t take advantage of the cloud’s redundancy and elasticity. Some vendors were trying to to take a a lift and shift approach. And what we, found were those that that was not an efficient approach. It’s, ended up, becoming exorbitantly expensive to run. And so out of that, CoralBay was born. Oh, very interesting. So Cloud Playout is quickly gaining favor because it’s letting broadcasters launch new channels faster, deliver more content in a more cost efficient way. What do you think is driving the shift towards cloud? Sorry. What do you think is driving the shift toward cloud based play out workflows and what obstacles do broadcasters face as they move away from legacy systems? That’s a good question. One of the benefits of cloud playout is it enables broadcasters to spin up and shut down resources. And so effectively, they’re only paying for for what they need. Launching channels in the past has been has has required a large investment in capital, making sure that they’ve got, you know, the servers, the air conditioning, the space. And with the advent of the cloud, that’s that’s no longer required. It’s a, you know, a matter of launching a channel in a few minutes, allocating your nodes. And then again, when you’re finished with it, tearing it back down. So yeah, you’ve only paying for what you use or what you need. And once you’ve done that, you’re finished. Yeah. I did see the infrastructure at work one time, your product CoralPlay is built on a cloud native architecture. It’s a suite of microservices, right? And I was impressed because you’re using technology in the correct way to address broadcasters issues, right? You know, why did you choose such a modern architecture and how does it benefit these broadcasters by leveraging these services rather than, you know, an instance or software that you run on a machine? Yeah. One of the benefits is that we use Kubernetes and Docker. And that provides what I see as three distinct advantages. One is it allows us to perform rolling upgrades. Secondly, as I mentioned just previously, is we can scale up and down, whether that be in terms of containers, adding new channels, resources. And then the third distinguishing point, I think, is is the self healing aspect. So Kubernetes has the the ability to monitor containers, container health. And so if if a container is upset or unhappy, you know, it can tear it back down, redeploy it, and you continue where you left off. And, you know, again, we’ve, you know, we we’ve made use of of Linux extensively, and that again brings with it many advantages in terms of like resource efficiency and things of that nature as well. Yeah. You mentioned, on demand scaling and it’s changing the game for broadcasters. Are you seeing a lot of broadcasters use that functionality to create pop up channels or event based channels for playout? Yes. Most certainly. What we see is that for for FAST channels, don’t need to potentially run for a long time, it may not make economic sense to invest in infrastructure on the ground and, you know, pay out for additional space or cooling or equipment and racking and things of that nature. So what I would say is the advent of cloud has transformed fast channels. Yeah. And we’ve been talking a lot about cloud. Some broadcasters are hesitant to take advantage of these new types of technologies. Is there a hybrid solution that you also provide for on prem or do you make it easier for them to get started with other services? Yes. So what I would say is we’re equally capable of running on the ground as we are in the cloud. And some customers, they already have what they need to be able to, you know, in terms of infrastructure to be able to run on the ground. And so it may not make sense for them to go to the cloud or they may have hopes of going to the cloud eventually, but wanna start on the ground. And the what I would say is at Coralbay, we sort of like to go on a journey with customers and help them help them transition. It also is the case that some of our customers will have legacy systems, traffic systems, MAM systems, and so you need to be able to interface with that. Again, we’re quite able to do that, support, you know, various subtitle formats. But customers may also want to have their system on the ground, but for newer channels deploy those to the card. And that’s probably a classic case where a hybrid approach makes, you know, good sense. Yeah. So are there customers that, you know what are what are some of the advantages of the hybrid approach versus, you know, a full cloud or for on prem? One of the benefits is that you can take advantage of both types of of hosting. So if you’re hosting on premises, it might be ideal for a channel that’s got a long life or a permanent channel, whereas the cloud would be ideal for pop up channels and it’s also ideal for disaster recovery. And so what you could do using our software is you can mix and match. So you can have some channels on premises and other channels in the cloud taking advantage of both. And it’s going to be the same user interface throughout. So operationally, it’s going to be the same, to to run it. And so you said that it’s the same user interface. So the user interface is what the user is gonna, you know, interface with. And does it matter if it’s running in the cloud or on prem? Or they don’t really see that? It makes no difference at all. It’s the same workflow, the same user interface. So operationally, it makes no difference at all. How does Coralbay make sure it plays nicely with legacy systems and processes? Yes, that’s a good question. So, we do interface with a lot of legacy sort of protocols or formats. We support a variety of different playlist formats. For example, Pebbles Marina or Pebble Neptune, also the Excel formats. We can also export to various different as run formats. Customers may have a variety of different media formats, and we support a whole range there. And that also extends to things like subtitle formats. There’s a lot of proprietary formats out there. So, again, on that front, we support a a wide variety. But, if a a customer has a a specific format that they need that they that we don’t currently support, we’re willing to to add support for that as well. Yeah. And at these trade shows, I don’t know if you guys have noticed, but there’s a ton of AI companies that wanna do transcriptions. Are you guys working with that? Are you doing your own? Like? We have recently done a project with the customer where they were using AI to generate subtitles. And we were taking those subtitles and inserting them into a stream, into a DBB compliance stream, all using the Zixi protocol. And this was using the EBU stand EBU time tech standard part three. So, yes, we’ve done we’ve have done some things with with AI. Yeah. And I forgot what which AI company, I think it was lingopal.ai that I saw there. I think they’re doing transcription and dubbing as well as audio. So are you guys also incorporating that into your workflows? We don’t do that ourselves, but we work with customers who do that and need to sort of in line the captions or the subtitles into the output stream. And is that something that is done more downstream of CoralPlay or is it can be incorporated into the production? Well, we are looking at incorporating technologies directly into the pipeline. So we have been exploring that. We don’t have a solution for that at the moment. Okay. But that can also be done upstream of of CoralPlay. Yeah. All right. So like, you know, switching gears a bit here, let’s talk a little bit about the ROI, right? You know, what, what are the key benefits for broadcasters and content owners seeing from the cloud play out in terms of like business agility? You know, does, does it reduce their CapEx spend? Is it a OpEx model? Well, I think that one of the big differences between launching a channel or launching channels on premises compared to the cloud is the CapEx spend. I mean, obviously if you’re doing it the old fashioned way, if you’re going to do it on premises, you’ve got to buy some hardware. You’ve got to find space for that. You’ve got to find some, some racks. You’ve got to then cable it all up. You then got to make sure you’ve got adequate air conditioning, and so on and so forth. And all is all the power costs as well. But when you’re running in the cloud, it’s very, very different. You don’t need that capital outlay apart from the initial professional services to do the setup and the and the testing. Then thereafter, it’s all subscription costs. So that kind of de risk things quite dramatically because if you’re noticing something new, you may not be sure that it is going to be a success. You haven’t spent all that money as capital outlay. So you can then shut it down if it doesn’t work and therefore you minimize the expenditure. Yeah. That’s, you know, that’s something that we don’t really think about, right? Like when you’re hosting this on prem, you have to deal with the power, the heating and the cooling. And if that goes, then you’re offline. I remember when I worked, used to have a portable AC system just for when the AC failed to kind of cool our systems down. So that’s something that, don’t really think about these days, everyone’s talking about moving to the cloud, but there’s significant infrastructure costs when you’re on prem. Absolutely. And real estate costs as well. Exactly. How do your customers measure the ROI when they move these workflows to the cloud? Well, it’s, it’s easier to measure. I mean, if you’re, doing it the old fashioned way on premises, then there’s an element of risk in terms of some equipment may fail and that may be unexpected. And then you’ve got to do something about that. So it’s a lot less predictable. Whereas if you’re running in the cloud, you know what your subscription costs are. And and your model is more aligned with the revenue model. Broadcasters get paid for what they do, and the income comes in and the expenses are all based on ongoing model. Yeah. Guess in the past, when you had infrastructure, it was a sunken cost and you’d have to figure out how to produce the content to make sure you make that money back. But now with the cloud, you know what the costs will be upfront, right? Yep. You know what the costs will be upfront and you can plan for that and cover that as you go. Yeah. And the industry is really changing FAST, not to use FAST again, but with shifting from on demand to FAST channels, it seems to be, a lot of people are moving in that direction, including sports. I can’t believe I saw that there’s a sports FAST channel, which is really interesting. You thought sports was just a live thing. How do you see these trends shaping the future of broadcast workflows, especially when it comes to playout? Yeah. We predict more FAST channels. I mean, it’s an opportunity for broadcasters to make more revenue from existing content. And so why wouldn’t they do it? So we expect to see the launch of more fast channels as as we move forward. But we also see, a move of, more traditional layout channels into the cloud as as time goes on. Do you think more of these traditional linear play out will move to the cloud, or do you think there there is a benefit to keeping them on prem? I think I think we’ll see more move to the cloud as people become more confident with the cloud. However, I don’t think we would go as far as to say that all channels will move to the cloud. I think there will be some that will remain on premises. It doesn’t always stack up if you’re doing a comparison between the cost, especially if it’s a channel that is a long term channel, a permanent channel, it may and if the infrastructure already exists and you have a long lease on a building, for example, it may make sense to continue that service, that channel on premises. And speaking about some linear channels remaining on prem, as these broadcasters are looking to move to the cloud, what advice would you guys give them to help them effectively move their product to the cloud and still maintain that high level of uptime and success? Yeah. That’s another good question. I think whenever somebody considers the cloud, I think they need to think of a few things. I think one of the things that I think is important to consider is what you would expect in terms of your success. What’s your criteria for success? Is it purely to save money? If it is, then you need to do an analysis over a good period of time to see if it’s gonna bring you that kind of return. But there are other benefits as well. I mean, we’ve just looked at some of them. We look at the benefit of the simplicity at which you can launch channels and also the simplicity at which you can decommission them as well. So those are the things that you need to consider. But there’s other factors as well. I think in choosing a provider for providing the software, you need to consider whether that provider is going to be around for long enough. You need to consider whether they are an innovative company that are going to be able to continue to develop the product. So as your needs change in the future, it’s also very important to choose somebody that’s gonna be there and support you in the future, I think. Yeah. And and you sorry. I I was just gonna say, I think it’s also for important for broadcasters to have personnel who have the skill set to work in the cloud. And it’s been a very different transition from, you know, how things were done in going back to the first question where things were done in a channel box all in the ground. You know, you need those skill sets of understanding Kubernetes and things of that to to be able to manage the, you know, manage the system and things like that. How how are you guys managing with the new skill set of broadcast engineering? We I you know, one of the other podcasts, that’s something we talked about quite a bit. Right? You know, the skill set of a broadcast engineer now has to change. They have to know some IT background. You know, how how are you guys helping to manage that process? Well, I mean, we have a fantastic support team and a team that also helps, you know, initially set up. So we offer services to do the initial setup. And as as I mentioned, you know, we go along on a journey with customers to help deploy the system and get it running and just use it day to day. You know, we provide training and things of that nature. So we try and make it as as easy as possible. But yes, broadcasters need to consider how that will impact them and the resources they will need as well. Yeah. And they typically would have SDI connections or ST 2110. How are you guys dealing with these other formats and getting it into the cloud? Or do you do you have a way or would you steer them back to using it on prem? Well, I think this is one of the ways in which Zixi has been revolutionary in that if you needing to get signals to and down from the cloud, you know, that has always been a problem. You know, you can’t default to just using TCP. You need to be able to monitor monitor it, check it for any errors, check it are those errors at the source? Are they were they being received? You need to be able to check bandwidth. You know, is it due to the connection? Things like that. So I think that’s where Zixi has really transformed it and and made our lives easier because it’s one less thing that we have to focus on. And I think the the the Zixi ecosystem is very mature and has made that sort of thing. You know, we have a tight integration with Zixi into our products to enable us to receive and send streams. Yeah. And that’s, yeah, been Yeah. Speaking of Zixi as a protocol, NDI was supposed to be the, you know, the production protocol to help you with all these types of things. Right? And and it only worked on local areas. Are you guys still seeing a lot of people use NDI for production? I think this is probably a very much on a personal question. I would encourage NDI use in terms of if you’re wanting to, like, monitor or preview. I wider than that, it it you know, people are willing to use it, but I think that’s yeah. Very much a a personal choice. Yeah. And recently, they came out with a new model where it’s all about paying for it. Do you have an opinion on how that’s gonna go? So we don’t use the advanced SDK. We’re using the basic SDK. And as far as I’m aware, that will not impact us. But, yes, it is it is interesting how that has changed from when it was originally conceived. And it’ll be interesting again if that changes further and how that may impact us. Yeah. Alright. Well, know, we did touch on AI a little bit and I like to finish my podcast with asking you guys, how do you use AI today in your day to day life? Good question. I use it extensively actually for coding. So for example, at the moment, I am writing a compositor to perform perform alpha blending for ten bits and twelve bits surfaces. And making extensive use of AVX2, AVX-512 twelve intrinsics. And the instruction sets for the for for for those for AVX-512 and 2 is quite comprehensive. And unless you do it every day, getting your head on the best way to achieve certain computations like averaging numbers and things like that is quite is quite complex. And I found AI to be really beneficial and helpful in breaking problems down into understanding how I can, you know, achieve a goal or calculate a certain sum or a function or things of that nature. So for my day to day job, it’s it’s it’s been really, really useful. Again, I don’t think it’s gonna replace a developer, but it’s it’s it’s almost like having an assistant to work alongside you is what I’d say. And Peter, do you use any AI? Occasionally, I I I use AI to help with documents to, improve the quality of documents. And, my my last cheeky question is, like, do you wanna share your last query on what you’ve asked chat GPT to do? No. No. I’ve nothing to say about it. My last query was how to do a podcast. Oh, okay. Oh, there you go. And did it give you useful information? It literally printed out these questions. Okay. That’s how I got these questions. These were not mine. Actually actually, on Friday nights, I was having a problem getting my I drive a manual car, and I was get having an issue getting it out of reverse and into first gear. And I wasn’t sure why that would why that could be. I wasn’t sure it was because of the linkage or the flywheel or some other major problem. And I asked ChatGPT, and it gave me some, like, possible things to try out to try narrow down the problem, which I thought was cool. It’s it’s funny because on the other podcast, it seems like people are now using AI as search rather than Google. Oh, okay. So instead, you know, you have a question and, you know, not so long ago, you would go on Google a question and you’d have to do a little bit of research. Yes. Now, as you said, it’s like an assistant. It goes out and does the research and it’s like, here’s the possible things. Yes. Because before you would probably have to go look several different forums. Yes. Try to figure out what’s going on. But it does that research for you. So it’s it’s interesting to see how, you know, some of those, things are coming into play. Yeah. So Peter, Gavin, thank you for joining us today. Thank you. We’ll see you next time on the podcast, I guess. Yep. Thank you.