Thought Leaders in Cloud Computing: Interview with John Michelsen, CTO of CA Technologies (Part 3)

Sramana Mitra: I know there is a lot of startup activity going on in that space. What are some interesting startups that you have seen, and what problems are they solving?

John Michelsen: I will start with the security space. You almost can’t talk somebody out of doing interesting innovation in security. My first prescription for going forward is you cannot do anything that is a risk to your security. You have to solve it before it becomes public if you don’t you threaten the business, the stock price, your customers’ loyalty to you and your brand. As fast as I want to go, as interesting and as bold as I want to be, and as much as I want to come out there and learn, I can’t do anything to affect my risk profile. Because of that, there will never be a reduction in the need for innovation in security. If you are a small company and are looking to get into something, you just can’t go wrong if you have interesting thoughts about security. I mentioned one around APIs just a few minutes ago.

SM: The APIs have proliferated and they continue to do so – everybody wants to be a platform now. The API security issue is very significant.

JM: That is right. Another of the trends I identified is that we are just at the beginning of what Google or any of those upstarts can now easily do in terms of digital assistance, the way we really want them to be – truly acting on our behalf, predicting our behavior, and offering us a better means to the end without being prescriptive.

That is the beginning of the agent-based movement, where our software agents are literal people on the web and who are going to interact. There are all kinds of new challenges. EasilyDo, for example, has a very interesting feature that tracks all the things it did on my behalf and then figures out how much time I would have likely taken if I did that on my own, eventually not taking almost any time at all. It is a fantastic thing.

But there is an identity and indemnification issue. We have to realize that if I am not really at the keyboard doing a particular activity at the time, what is the recourse for me doing bad things? Am I really the person who is represented by the software that is doing all of that stuff? It is fantastic that we need to solve these problems, because it is a fantastic future we can create for ourselves where some of these interesting tasks can be handled for us. At the same time, we have to figure out how to do that in a world of safety. I love that problem because it sounds really hard. It is certainly something that smart people can focus on and there will be lots of ways to solve it, and someone will win and be rewarded for that victory.

But there are also countless other opportunities. We have seen Java as a development language dominate enterprise applications for a long time. I think that time has come to an end. It has already become a very distributed and multilingual presentation layer on mobile devices, applications, websites and presentations. I think that will only continue to push down into the application layer where it is very common to hear Python, Ruby, etc. I think those will continue to expand. While five years ago I would have said, “Don’t go build a development tool, don’t go build in the language world, because that is locked up by Java” – that is obviously not the case anymore and interesting things can be done there.

SM: I also think the platform as a service trend has also changed that.

JM: That is right. I think there are interesting things going on there. CA itself has a significant investment – which happens to be my responsibility – to bring together the various platform and core technologies that we have been undertaking for many years now and to bring that to a pass level and for that to be the underpinnings of all of our products, to enable better usability, interoperability and massive cloud scale. We have a legacy of products that we have to put on a rapid journey, and we have to essentially re-engineer the bottom of all of these products to make them able to make that journey.

Platform level technologies – it’s an interesting thing. Who would have thought Hadoop, Cassandra or NoSQL would have had a legitimate shot at unseating Oracle or typical messaging systems? We wouldn’t have thought that just a few years ago. I am not a big believer in the current technologies we have for big data. I think Hadoop and others are a great shot out of the gate, but there is a lot of maturity that we can reach there. Big data is much bigger than map producing. I think there is more innovation right now in analytics than in core big data collection and refinement.

This segment is part 3 in the series : Thought Leaders in Cloud Computing: Interview with John Michelsen, CTO of CA Technologies
1 2 3 4 5