"Everybody is interested in AI but not many companies really have AI in production",
says Media Distillery chief executive Roland Sars.
Can you outline the present composition of Media Distillery’s work and how operations have changed during your three years at the company?
The team has grown tremendously in the past three years. When I joined [as CEO and ‘late founder’], we were with 3 people and we did not have any employees yet. Now the team is about 30 people. We hired a lot, mostly on the technology side, but we also grew the team in terms of sales and marketing. In the initial years, the company was focused solely on media monitoring; that is providing companies, brands and organisations with information about when they have been mentioned on radio and TV, and send a real-time alert to them. About three years ago when I joined Media Distillery, we also started to look at other areas, and the broadcast and TV areas seemed quite relevant for our technology.
Firstly, there is a lot of video content created and distributed on a daily basis in this industry, and we can add the most value when we process a lot of data. At the moment we do more than 20,000 hours of video on a daily basis.
Secondly, these companies, mostly TV operators, have limited understanding of the content they distribute such as: when the programme starts, when commercial breaks start, what topics are being talked about, and which people are actually visible in TV programmes. With personalisation, recommendation and search being more and more important, these metadata are quite relevant.
Lastly, the media industry is changing drastically with new OTT players entering the market, [including] Netflix, YouTube and Hulu. The traditional TV operators need to innovate and change the way that they see content and what they do with it.
All in all, not only Media Distillery has changed a lot in the past three years, but also the environment and the industry with which we are working.
Media Distillery’s flagship is the Deep Content Understanding technology. In what primary ways can this be advantageous to media companies?
There are a few ways in which it can be an advantage for media companies. More and more consumers want to have personalised content; for example, relevant search results, recommendations or a personalised playlist with videos to someone’s taste.
The users know what type of content they want, and they expect to have it presented in a nice way with a great user experience and smart user interface. While the market is changing nowadays, TV operators need to know what their content is about and understanding video content is what Media Distillery delivers. We create the next generation metadata which helps them to really understand what is inside of their content and can enable them to offer it in a better way with an advanced user experience.
AI is very much a hot-button topic at this point. Where do you think we are on the trajectory in terms of understanding within the industry about the true potential of AI?
My experience is that everybody is interested in AI; some companies are applying it, some are doing internal research about how to do it, and some are testing, but not many media companies really have AI in production already.
In our latest blog, our machine learning expert explains our view on this well. For most of the companies interested in AI, it is still in its infancy with testing. I think that it is also where the value and the power of Media Distillery lies. We can really hear and understand the market, where challenges are, and we can apply our technology [accordingly].
In what ways is the IBC exhibition important to your status as an AI innovator and in terms of meeting potential new customers?
IBC every year is very important to us and next door as we are based in Amsterdam.
The global industry comes together for five days, which provides a combination of three opportunities: first of all, we can organise meetings with existing customers to share our updates and latest technology innovations; secondly, we can have a lot of meetings with international prospects, which saves on our costs; and lastly, it is a good podium from which to make a new product and partnership announcements. For these reasons, IBC is for us the most important event of the year.
Can you briefly outline your plans for IBC2019 with a specific focus on any new launches and project announcements?
We are planning to announce a new partnership with a large European TV operator around IBC that will be launched in our AI platform. As well as this, we will be focusing on in-video search capability with two applications for TV operators.
The first of these is Snackable Content, which is an in-video search solution enabling TV operators to enhance the consumer experience in their replay environments and provide the perfect VOD user experience.
TV operators can deliver better search results and recommendations based on enriched metadata that we create. Our technology distills the video content based on a consumer’s favourite person, topic or interest. This metadata can be used to tailor video recommendations for every individual consumer.
Our technology can also be used for advanced advertising, with the advertising ecosystem now moving from broadcasters to TV operators. The TV operator is dynamically inserting ads based on their demographic knowledge about the user, but they don’t know what exact content is watched by their audience. So when you match these two data types – demographics and rich metadata – you have all the information to create a great personal profile of a user or household.
With this advertising profile, advertisers [have the opportunity to] better target their audience, which makes advertising more relevant and means brands are willing to pay more. It is a new revenue-generating business for TV operators, so we are investing a lot of effort into this development. It’s a way in which we can help them make these business cases more useful.
July 2, 2019
