By Sramana Mitra and guest author Shaloo Shalini
SM: Your point regarding the uncertainty associated with integration cost savings in the cloud is well taken. I just wanted to make sure that you are not biased by the fact that integration is Mahindra Satyam’s bread and butter and you do not want to see that costs go down with the cloud.
RN: Not at all. I know that to some extent we may have to cannibalize our own business model. I am absolutely sure that we have cannibalized some of our own revenue stream, and that is why I said systems integrators will have to think differently. We as system integrators need to think differently. So that is not making me biased; the only caution that have is on the maturity curve.
SM: Yes, absolutely.
What you are emphasizing is that we don’t have enough data from cloud adoption experiences yet. We don’t know exactly what is the real integration cost and associated cost savings in terms of what is really happening when the rubber meets the road. I think that data is going to come together in five years as we have enough of this infrastructure moving to the cloud; that’s when we are going to have real data and that is when we are going to know the answer to this question.
My vision is that if we are able to evolve the standards built on fundamentals of service-oriented architecture (SOA) rapidly enough, we will see some of this happening. I believe this will happen because companies like IBM and HP are the ones that will make sure more functions move to the cloud because it is very interesting for their business as well. I truly believe that eventually integration will be offered as a service out of the cloud. It will be possible to connect the dots without too much effort and bring the value of aggregated information to your business processes in the enterprise. Integration as a Service will also be possible on the cloud. That is my firm belief. The only caution I have is, it will happen once cloud computing reaches level 4 or 4½ of that maturity curve .
SM: What you are saying is that from a CIO’s point of view, you need to design the entire infrastructure, the entire data infrastructure, and I guess the entire IT infrastructure in a SOA-compatible mode so that you can elegantly draw from that architecture the integration elements.
RN: You articulated it much better than I could have. Yes.
SM: You used the word “standard”; what are you seeing in the market in terms of standardization? What has caught your attention in terms of standardization in cloud computing?
RN: So far, I don’t have much to say on standardization in the clouds, but I can feel the buzz happening. I was just reading a blog a couple of days ago and I read that, for example, IBM is beginning to make some noise about SOA standards, what they means, how they can elevate them to a consortium and so on. I think that is nice, but I haven’t seen enough of it. It will happen very rapidly, trust me! Standards at some point will almost be forced because otherwise cloud computing will not go anywhere.
SM: Where do you expect that standardization to come? Is it is going to be data model, and is there going to be a standard vocabulary of CRM data that everybody is will follow and the analytic engines will work off those? What is the nature of standardization in the cloud?
RN: I will give you an analogy. Are you familiar with SCOR?
RN: It is a nonprofit organization; it stands for Supply Chain Operations Reference model.
[Note to readers: Refer to this related discussion about SCOR-based “super specs” in automobile after market – created by Gcommerce, it is the backbone of their cloud computing deployment and use of cloud in the automobile aftermarket supply chain ecosystem.]
RN: What happened was, SCOR was born in 2005, and today it has attained probably level 4 on the maturity curve. I’m just arbitrarily using that level so you can get a perspective. Now, this organization realized that part of the challenges in the supply chain field, and it had nothing to do with IT by the way, was purely concerning the supply chain as a process. So, part of the challenge of the chain process is that the taxonomy was all over the map. The ability for anybody to measure performance, report on service-level agreements (SLAs), and sign contracts that are based on SLAs was all over the map. So they embarked on the journey and said, We are going to break this down, we are going to take it industry by industry and create a taxonomy that everybody begins to understand. Now they have a very large membership. I think they have more than 6,000 members.
SM: I see.
RN: And if you look at the evolution over the past 4½ years, it’s phenomenal because now they have taxonomy by industry, and it is very well orchestrated and people have better control over supply chain processes. If you go to a company of any repute and say you’re a supply chain consultant, the first thing someone will ask you is, Do you understand SCOR and can you help me deploy it? So, that is the same principle in which cloud standards will have to evolve, and those standards will have to be the way we create the data taxonomy, event taxonomy and process handshake taxonomy. These taxonomies have to evolve; otherwise, it is not going to be possible for us to leverage cloud capability.
SM: If you look at what is happening in the world, take the example of the wholesale business Now, with e-commerce – I was just looking at some numbers – 20,000 e-commerce companies are starting every week in the United States alone. We haven’t looked at China, we haven’t looked at India, we haven’t looked at any of this. Now, think about a manufacturer who wants to sell to this far more fragmented, far larger set of retailers. For every single category a manufacturer now wants to sell to, there are an infinitely larger number of retailers that they could be working with unless they are some parts of taxonomy that plugs in to these retailers in an elegant way. This is not going to be a manageable situation.
RN: Exactly! The other thing that is going to happen is that the cloud will recast conventional programming. I am no crystal ball reader, but I can tell you that programming will be passé in the next five years or maybe ten years because everything is going to be orchestrated through events. Event A happens and triggers event B and so on. There is going to be a shift in what I would call as not programming but “event gramming.” This could be an interesting paradigm shift that I think IT companies have to get ready for.
SM: I don’t think it will happen in five years!
RN: Okay ten, maybe after I retire.
SM: The industry moves a lot slower; the kind of standardization you are talking about that would lead to “event gramming” as you are envisioning requires an orchestration of so many different players at such a massive scale. That kind of coordination, that kind of thinking up does not happen. I mean, look how long salesforce.com started, way back in 1999, look how long it has taken for cloud computing to gain adoption, and now eleven years later we are still level 1 of the adoption process, or level 2 at best?
RN: You have just poured cold water on my optimistic vision!
SM: Yes, I did.
RN: I’m just kidding! No, you are right! You are being practical.