SVG Europe Sit-Down: Cloudian’s Jon Toor reflects on AI and the next generation of archiving

Jon Toor, Chief Marketing Officer, Cloudian

Cloudian is a Silicon Valley-based file and object storage company specialising in S3 API storage systems. By contrast to proprietary, and often expensive, object storage products which came before it, Cloudian’s technology allows all sizes and types of users to realise the benefits of object storage in their own data centres. Our Sit Down with Chief Marketing Officer, Jon Toor, began with a few thoughts about NAB 2018…

What were visitors to your booth at NAB most interested in discussing with you?

We heard about the challenges of managing growing media archives – how to deal with exploding capacity demands, how to find media faster for rapid re-purposing, and how to easily integrate media storage with the cloud.

So, firstly, scale. 4K media is now commonplace and 8K media is well on its way – users know that capacity growth will continue unabated and traditional storage solutions are not keeping up. Disk solutions are expensive; tape is slow and unreliable. What they need is cost-efficient storage that scales on-demand.

Secondly, searching for media is just as challenging as storing it. A database works great with structured data but has limitations when dealing with unstructured data. Many visitors wanted to learn more about metadata, which can help solve this business concern. Metadata is ‘data about data,’ which can be information on the contents of the data. With metadata, you can search your entire archive in seconds using Google-like tools, quickly finding specific clips that contain a certain player, for example.

Lastly, media professionals know that the cloud will play a role in media distribution, or in services like transcoding and disaster recovery. The question is how best to integrate the cloud with on-premises storage. The two environments must be able to work seamlessly together, without middleware shims that add cost and complexity. Our visitors wanted to know how our solutions help to merge these two worlds.

What are your biggest challenges as you move forward after the event?

We heard from many new prospects at the event, many with new use cases – we also have received a lot of new partner interest, too. We are now ensuring our resources are perfectly positioned to meet these expanding demands. Managing that growth is our focus.

How do you see Artificial Intelligence and Machine Learning (AI and ML) impacting your solutions?

AI/ML offers tremendous potential to make media search vastly more powerful than it is now. By enriching metadata, AI/ML tools can add new information – such as player names, or venue names, or weather conditions – to help producers quickly find the exact clip they want. You can go back in time, drawing on information from images or audio transcriptions, to create a sort of instant replay forever.

What questions should a potential customer ask of a supplier to ensure the right package is purchased?

You want something that fits your budget today and can grow limitlessly in the future. So, ask about system scale, that is, not only how large it goes, but how small, as well.

Cloudian Hyperfile 3: Cloudian HyperFile delivers limitlessly scalable enterprise file services, ‘on-prem’.

Data protection is a massive deal now. Companies, customers, consumers, everyone is thinking about data and how organisations are actively protecting it. Protecting a multi-petabyte archive is different from backing up your application servers. So, you should be asking suppliers about data protection: How is the data protected, and do I have to buy any additional features or software licenses to get that protection?

The cloud is increasingly a part of the equation, so ask about cloud integration. Can I integrate with Amazon/Google/ Microsoft out of the box, or does it require extra software? Ask the vendor to show you that media moved to these cloud providers is accessible by cloud-resident applications. With some systems, the media is stored in a proprietary format, which limits your options. These are important things to know from the offset.

Apart from the obvious increase in file size, does archiving 4K – and in the future 8K – require any new thinking on the part of archive solution providers?

Beyond the higher data rates of 4K media, the larger issue is that cameras are now rolling all the time and producers want access to everything that was shot. That requires greater capacity and a better way of cataloguing and locating media.

Additionally, media is being re-purposed after the fact, either to create entirely new content or to rework existing programming for new formats. Either way, archives are no longer static. We used to write a tape, put it on the shelf, and rarely touch it. No more. Archives now need to be dynamic, and easy to search.

What should we look for in the next generation of archive solutions?

There are massive strides and achievements being made in archiving solutions, with advances revolving around four areas – limitless capacity, rich metadata and Google-like search, built-in data protection and disaster recovery and Cloud integration.

Do you have a European broadcasting case study you can share?

None that we can share in public right now, but please check with us at IBC in September for the latest news!