Discovery engineers “predictable manufacturing” supply chain as it makes its way onto different screens, cloud platforms and sports programming.
http://www.csimagazine.com/csi/Discovry-cloudi-its-supply-chain.php
“We are no longer a TV company,” declared John Honeycutt, CTO/CIO - Discovery Communications in an IBC keynote. “So much of what we do is about re-engineering supply chains to reach out to many different platforms.”
Discovery has moved from a purely factual lifestyle business to being just 140-odd days away from broadcasting the Olympic Games on Eurosport.
“Cloudifying the supply chain means preparing for multiplatform exploitation,” said Honeycutt. “It’s about saying where are the assets, where do they go and how do I get them there as efficiently as possible.”
Discovery has over 600 content suppliers so working out how to scale those workflows is a key part of Honeycutt’s agenda.
“The goal is make the costs predictable,” he said. “By reaching further up into our supply chain the more we can understand about the intent for each piece of content – is it VoD or live – then the better we can design automation downstream. We want to be working in a predictable manufacturing environment, not an unpredictable one.”
He added, “Doing that means, for example, I can scale transcoding engines and ultimately maybe we’ll help create a marketplace for this. If I have ‘X’ amount of hours to transcode how much will you charge me?”
Moving to cloud is all or nothing as far as Discovery is concerned. “It’s very difficult to put a toe in the water. The economic model of moving content up and down is where [cloud vendors] are making their money so you have to have a holistic strategy and that starts with controlling content. The further upstream (to content suppliers) you reach, the more complex it all becomes but that is where the efficiency is.
“The goal is to put efficiency into these kinds of processes at a centralized level while retaining localization and personalization. We have to avoid the spaghetti of complexity.”
Balan Nair, Executive Vice President & Chief Technology Officer - Liberty Global, highlighted the huge disruption in technology and in business models adding that the demographic shift is the widest it have ever been between generations.
“The younger generation’s ability to adapt to technology is unbelievably fast,” Nair said. “Then, commercially, you have irrational competitors with a different cost structure to you. The cost of getting content to customers is more than $10 a month for companies like us – so how do we deal with that?
One way is with machine learning. Nair said AI was already pervasive throughout Liberty, mostly in customer interactions such as with chatbots. “Our new voice remote has a natural language processor which is ML-based. Since we have services in so many different languages you need ML technology to make it work.”
He added that about 15% of Liberty Global’s processes and systems incorporated AI today, a figure he predicted will rise to 50% by 2022 and higher still beyond that.
News Corp’s Global CISO and Deputy CTO, Latha Maripuri agreed: “AI is here and it’s more and more critical for business whether that’s how you manage the enterprise, or security, or to help create and collaborate on content.”
Was anyone concerned about placing so much of their infrastructure into the hands of essentially one of three major cloud providers (AWS, Azure, Google)?
Nair replied; “Certainly by building your own virtualized network you can control your own destiny and control your own data. You can control latency and be as close to the customer as possible, plus if you build yourself you can capitalise on that infrastructure otherwise all your operations costs goes to Microsoft or Amazon.”
But while AWS had experienced outages – “one which took us all by surprise earlier this year” Nair said – “any of our networks have had more outages. You should resist teams trying to convince you to build your own cloud. You will be left with a lot of cost.”
No comments:
Post a Comment