You’ve recently started showing people the first demos of Aurasma, but I’ve seen comments on the Web from people saying that you can’t possibly have built this yourself so quickly.
It actually took about a billion dollars [of investment]. The core of Aurasma is the ability for a mobile phone to recognise half a million objects and do that under real-world conditions – that is cutting-edge, PhD-type technology. We’ve done it because we have taken our IDOL [Intelligent Data Operating Layer] platform, which is what the last billion dollars of R&D have been invested in, and spent a year with a large number of very good programmers getting that to run on a phone footprint and also understand the 3D model. It’s absolutely fundamental technology and very difficult to do.
There are about 100 organisations starting to work with Aurasma. What are some of the use cases?
You have film studios taking characters from their upcoming movies and putting them around the major cities so you can walk round New York and meet them [they become animated if you point your iPhone or Android smartphone at their image on a billboard, for example]. We’ve got games companies where you make the games location-based – so you are physically going round places and doing things as part of the game – and museums where the exhibits come alive and tell you about themselves.
We’ve also got one around missing children – it’s OK seeing a picture of a child, but what if that gives you a video? It’s much more emotive and gives you a lot more information. There are also travel guides, where you can walk around ancient Rome.
Obviously, advertisers are doing a lot of stuff. For example, hold the phone in front of cereal packets and get a lot more information, recipes and so on. On top of that you have user-generated stuff – it’s amazing what they come up with, completely unexpected things that appeal to their sub-culture.
Why did you buy the digital archiving and online backup assets from Iron Mountain and when did you become interested?
Iron Mountain was one of the last players in archiving when we entered that market in around 2007. IDOL gave us a number of advantages – we’ve got up to 19 petabytes of information – but because of the amount of information it takes time for customers to switch. Iron Mountain had got up to about 6 petabytes – that was the relative position – and in 2009 we started talking to their CEO about a deal. Now we have been able to conclude it.
You launched your Structured Probabilistic Engine (SPE) in September 2009, which you said helps to give structured databases shades of grey instead of only searching in black and white. How have sales gone?
It’s been one of the fastest-growing areas of the company. In fact, I think it is the fastest growing. It’s growing at 60% or 70% quarter-on-quarter at the moment. A lot of sales are through OEMs where we license it to players in the structured market, and they go and write the next version of their products, and that’s about two years so you’ll see a lot of those products coming to market soon.
How behind cloud are you? Would you like Autonomy to be seen as a big player in cloud?
Well, Autonomy has been doing cloud since before it was called cloud; we’ve been doing it for 10 years. We have 25 petabytes of customer data in the cloud – I think that makes us the largest cloud provider. You could have a debate about Amazon or salesforce.com, but the difference is what we have in our cloud, which is all the day-to-day interactions inside a company. That’s the most valuable data.
Listen to the in-depth audio podcast with Mike Lynch at www.cbronline.tv.