Ahead in the Cloud: A New Era of Cloud with Anant Adya

Chad Watt: Welcome to Ahead in the Cloud, where business leaders share what they've learned on their cloud journey. I'm Chad Watt, Infosys Knowledge Institute researcher and writer. Today, we're doing something different here on Ahead in the Cloud, it's September 2023, and Infosys has just released our Cloud Radar 2023 research report. I had the privilege of sitting today with our top cloud leader, Anant Adya, to discuss the work. Anant is the EVP for Infosys Cobalt. Welcome to Ahead in the Cloud, Anant.

Anant Adya: Thank you, Chad. Thank you for the opportunity.

Chad Watt: This report is the second time we've studied corporate cloud deeply, and its results come from a survey of 2,500 cloud executives and interviews with more than 50 business leaders and technology experts. What's new in the cloud Anant?

Anant Adya: Everything is a new in cloud. When we studied this report in 2021, Chad, if you remember, it was all about save and optimize. And customers who were consuming cloud were looking at cloud mostly from saving and optimization standpoint, but now things have changed and cloud is being looked and leveraged for grow and disrupt, and that's what is new and that makes it extremely different. And the way customers are leveraging cloud, I would say, is very different and new.

Chad Watt: Talk a little bit more about cloud for integrating acquisitions because the nature of cloud strikes me as that ought to be something cloud is very good for your connecting systems across distances, across different platforms, cloud should deliver on that, right?

Anant Adya: Well, absolutely. To do that right, we have to go back to the basics a little bit. Unless you standardize and rationalize your IT environment. Integrating new capabilities or integrating new companies becomes difficult. So we do have customers today, for example, we have a healthcare customer who actually leverages onboarding a new acquired entity, much, much more efficiently than the way they used to do it before cloud. And the only reason they have been able to do that is because they have brought in a whole lot of standardization and whole lot of rationalization in their environment. So when you use reference architectures, business blueprints, and you actually do not reinvent the wheel every time you do it, it becomes much more easy. And that's the area that customers are learning and it is taking some time for them to do it.

Chad Watt: How much of that is technology process and how much of that is business process?

Anant Adya: So I would say business process is a big part of that. Technology is very easy, right? Because technology is always a tool that you use to solve a problem once you find the problem. So I would say it's more about business process, and once you rationalize and standardize that, it becomes much more easy to solve it using technology like cloud.

Chad Watt: Let's talk about cloud spending. What did Cloud Radar discover about cloud spending?

Anant Adya: Well, I would say more companies will increase their cloud spending. That is the bottom line that we found. Two out of the three did it last year, and four of the five will continue to do in years ahead, right? I mean, that's essentially what the Cloud Radar has found out.

Chad Watt: Why are people still spending more in the cloud?

Anant Adya: More and more customers are leveraging cloud to grow their revenues, to create better products and solutions for their customers, to expand into new markets. And in general, they're trying to see how they can grow and disrupt, that's essentially one reason. And when you go to grow and disrupt, you have to be very fast to the market. It's all about faster time to market, and faster time to market can be achieved when you have the power of cloud, whether it is about getting huge amount of infrastructure to fail fast, learn fast, it is about getting compute storage, it is getting technology in the cloud, which is elastic, scalable, highly available and available globally, you will need cloud. And that's where the cloud spending is increasing. Number two reason is the gen AI, which is the latest and the hottest word in the market.

Chad Watt: Gen AI, being generative AI.

Anant Adya: Generative AI, yeah. And essentially a lot of customers who want to use either small data sets or large data sets to train AI models require huge compute capacity, and you will not have the money and time to sort of buy that compute capacity, set up your own infrastructure, build your data centers, and for those kind of initiatives, you need cloud. And that's the second reason why the cloud spending is increasing in the big way.

Chad Watt: All right, cloud spending is up. You've given me some inputs on how, but are they spending it with the same people or are they spending it with other people?

Anant Adya: Although the number of cloud players or cloud providers in the enterprises have gone up in a big way. I think when we started talking about cloud many, many years back, it was one provider doing everything. Then there was this competitive advantage that the enterprises had to get, and so they brought in another player. But today we talk about three to four players in the account or in the enterprise who are actually delivering cloud work.

Chad Watt: Would it be possible for a big business, one of these global 2000 businesses to go back and really just roll back to one single cloud provider?

Anant Adya: What we have seen from the report is one in five companies do that. But see, when you talk about cloud, it is not just about infrastructure as a service. It has IaaS, PaaS and SaaS.

Chad Watt: PaaS and SaaS being platform.

Anant Adya: Platform as a service and software as a service. So when we do that, you need more and more cloud providers to be present, and that's essentially why it is more than one player.

Chad Watt: Okay. So here we are, cloud works well, it's used for more and more things across the enterprise. Companies are spending more, they're using more cloud providers. That sounds good. Are there some concerns emerging around the edge of this?

Anant Adya: I think utilization is a big issue. I think one of the things that we have seen, and there are multiple reports, there used to be these consumption commits that enterprises used to make to the hyperscalers. And broadly speaking, when we talk to our customers and when we see reports, we believe that 47% of that spend has been sort of, I would say leveraged or utilized. So the utilization of the consumption commits have actually stayed less than 50% over the last few years.

Chad Watt: So companies aren't using all the cloud they've committed to?

Anant Adya: No, they're not. They're not. And that is a big, big reason for concern, that even the hyperscalers and GSIs like us are having in terms of why that consumption is not going up.

Chad Watt: So it's a problem for the company buying the cloud. It's a problem for the provider and a problem for their technology partners. Why is that a problem for the company that committed to buying the cloud? I don't have to pay for something I haven't used, do I?

Anant Adya: Yeah. So I think essentially they will have to, because these are consumption commits. These are contractual commitments that you have made to the hyperscalers and you really cannot get out of those consumption commits. And the reason why the commitments have not been fulfilled or the commitments have not increased beyond 50% is predominantly because the way enterprises were consuming cloud. And essentially, as I said, all those customers who just decide to consume cloud for infrastructure as a service, where they said, "We want capacity in the cloud." And they were leveraging the capacity just for infrastructure have not been so successful in terms of using cloud. Because if you use cloud just for infrastructure, it is expensive. And sooner or later your CFO is going to come and ask you why are you spending so much money on cloud? And that's one big reason. And in fact, there is a public report that you would've seen. This is on public domain.

Sabre made a 2 billion commitment with a hyperscaler, and 18 months into it, they had only consumed 10 million. And this is a classic example in terms of why you need to have a right strategy in terms of where you're going to leverage cloud, how you're going to leverage cloud, and what kind of commitments need to be made around that. I think that's essentially what has been missing.

Chad Watt: How do you, as a corporate leader and advisor to a corporate leader, predict what cloud that company's going to need 18 months, hence, three years, hence?

Anant Adya: Well, absolutely. That's why we have this particular service called Cloud FinOps. Essentially, Cloud FinOps is all about three things. I mean, one is looking at what you have today, because it's very important for clients to know where they're spending on cloud.

Chad Watt: Wait. Do they really know what they're spending? I mean-

Anant Adya: Yeah, a lot of them don't, because everybody was swiping their credit cards, the testing team was buying, the development team was buying, the business was buying, and you know what, IT was buying. So number one is to get an inventory and get the knowledge of what you have. That's number one. Number two, now that you have details in terms of where you are spending on cloud, you need to now look at how can I optimize that cloud spend, right? That's the second aspect of Cloud FinOps. And the third aspect is do a forecast in terms of what will you require in terms of your cloud three months, six months, 12 months down the line, right? Because that's very important. So if you follow a very robust methodology of Cloud FinOps, you'll be in far better control in terms of your utilization.

Chad Watt: You mentioned FinOps, and I like where you described that and know what you have optimize and then forecast ahead. If I had to ask you, out of 100 big businesses, how many have a good understanding of what you described there and have put it in place? How well known is FinOps, and how well adopted is FinOps?

Anant Adya: It's pinching the budgets of our customers, so it is very well known, and clients also realize that they need to do something about the cloud spend. Because cloud is not cheap, cloud is expensive. And unlike in the past where hyperscalers used to reduce the cost of compute and storage in the cloud, recently the projects have gone up. So Cloud FinOps is pretty much known to everybody. Now, what they need to do is to put a very robust methodology for Cloud FinOps. You can't take an approach of taking a very tactical approach to Cloud FinOps. It has to be very strategic approach to Cloud FinOps. And that's what is something that we are educating the customers, and that's what is missing today.

Chad Watt: Cloud is where companies do business. It's where growth disruption transformation happen. If you're a growing company, it's inevitable you're going to spend more in cloud to some degree, but how do you get a good handle on how much more you should spend? If we're growing at 20%, how much should our cloud spending be growing?

Anant Adya: Essentially, the correlation is all about embedding the business case in the project that you're doing. As long as you can get the better return on investment using technology and spending money on technology, it does not really matter. But when the amount of money you are spending on the digital transformation project or any growth project from a technology standpoint is much more than the returns that you're getting, that's where the problem is. That's why there is no direct correlation as such in terms of how much money you should be spending on a project. But correlation is very simple. It has to be the business case.

Chad Watt: Got it. Correlated to the business case. I like that. We've talked about cost. Let's talk about security and governance. What are the emerging complexities around security and governance in terms of cloud?

Anant Adya: One is, I think if you look at our Cloud Radar, I mean when we interviewed our clients, we got to that 43% of the companies have very lax policies around how they manage security in the cloud. So that is number one. Number two, 1/3 of the companies have IT only or business, right? In charge of cloud purchasing, deploying, retiring, and compliance, which means many more or the 2/3 of them have cloud, which is being part by everybody in the company. So there is no control on how cloud is being procured, deployed and managed. So these are the two issues that lead to issues with respect to security, governance and compliance.

Chad Watt: So we're kind of entering a new era of cloud, and that requires a new approach. That's what the report tells us. Cloud keeps getting better and it keeps growing more complex. And Cloud Radar 2023 argues companies need a new approach to how they manage cloud. Give us some advice. How can you do that?

Anant Adya: Cloud has moved from just being a tech initiative to a business initiative. It is no longer about using cloud only for technology related aspects, which means that the buying centers are not just the CIs and the CTOs, but there are other stakeholders in the company that use cloud. And when you have this particular option available with you, if you don't evolve the way you manage cloud, you are at a risk of losing big time, right? I mean, that's essentially one of the issues that we see. I think we recommend three step approach to the way you should manage the whole cloud thing. So one is master, monitoring and prediction. I think this is very, very critical. You have to monitor the cloud, you have to predict what you need in the cloud and you need to master that particular thing, right? Number two is embedding the business gracing cloud, everybody embarks on a journey of digital transformation and they say cloud is going to be table stakes and we are going to use cloud, but how do you make sure you embed business case in the cloud? I think that's the second aspect of it. And the third one is adopt a value centric cloud operating model. That is a very important thing, in the sense what is the value that the cloud operating model is going to deliver.

Chad Watt: Fall 2023 will be about one year from ChatGPT's big live open trial splash when everyone could have artificial intelligence on their mobile device. There are some products out there doing some simple things, but I'd argue we've hit a trough in where generative AI is. Do you think generative AI is still going to be a game changing technology?

Anant Adya: Well, absolutely. We absolutely believe that gen AI will be one of the disruptions that is pretty much going to become mainstream for every enterprise. And we already started seeing a lot of that impact in the way customer experiences are being delivered, whether it is call center, contact center technology, or whether it is about how huge sets of documentation is being used and how knowledge is being used and those kinds of things. There are very, very many use cases that we are already seeing. And essentially the way cloud will play a big role in AI. I have two predictions on that, right? And predictions or hypothesis as we can call it. Number one is most of the cloud consumption in the future is going to be driven by AI projects. And the reason is very simple. Customers will use pre-trained AI models. They will create their own AI training models from scratch, and they will require a lot of compute and storage to do that training. And that compute is not available in the data center, it is in the cloud. So if you want to train, deploy, engineer and manage these large AI models, you'll need capacity. And that capacity is something that you can get from the hyperscalers in the cloud.

Chad Watt: AI uses compute at an exponential rate. Are there some ways you can control for that? I mean, it's different from me saving my documents into the cloud.

Anant Adya: Great question. In fact, we were just discussing about that with one of the clients, and it is not necessary that you have to start from scratch. The CTOs or whoever is looking at AI models has to keep this in mind that you don't have to start everything from scratch. There are pre-trained AI models that are available, I would say in the public domain, which can use and customize and use them for your gen AI project. So that's number one. Number two, if there are complex business problems that you want to solve using AI, you don't have to go for those million data sets and billion data sets, right? You can use small data sets, which can still solve your complex business problems. And lastly, if you still want to start from scratch, you have the hyperscalers who have these GPU capacities that they can bring from NVIDIA and a couple of other companies which can help you to do that, right? So essentially, my recommendation would be if you want to contain the cost of cloud or consumption for AI, one, use pre-trained AI models, number two, use small data sets for training, and lastly, if you still have to use start from scratch, there are the hyperscalers who will provide you capacity on demand as a service without spending tons of money for CapEx.

Chad Watt: So let's talk about industry specific cloud. Cloud Radar found that cloud was rather the same industry to industry in terms of productivity and business processes. It works quite well for those things, and that the industry cloud solutions are in a fairly early stage. We're going to continue studying this some more, but let me ask you, assume that this is the early days for industry clouds. Where do you see industry cloud developing most rapidly and why?

Anant Adya: So we in fact, at Infosys, right, as part of Infosys Cobalt, we did launch financial services industry cloud very early on, and we had specific industry use cases that we brought in specifically for retail banking, for fraud detection or monitoring, for reducing the anomalies in banking transactions using risk monitoring and those kinds of things. So those are some specific use cases. We also launched CPG retail industry cloud. So there is a lot of things happening in the CPG retail side with the supply chain, with inventory, with how do you manage pricing optimization and all those kinds of things. Actually, there are certain industry use cases there as well. Lastly, in May, we launched airline industry cloud focused on airlines and how they can bring some of the use cases that are catering to the travelers like us, I mean, who are always on the road. So I would say there's a very distinct capability that industry clouds bring, which is all about business agility. If you want to focus on IT agility and IT efficiency, it's all about using infrastructure charges service, scalable compute, CapEx to OpEx, legacy modernization, those kinds of things. But when we talk about industry cloud, it is about bringing business agility, and that's the value that actually industry cloud spike.

Chad Watt: Can you shed some color on how companies spending an investment in cloud has changed over the past two years, and forecast ahead, what it's going to look like in 2024 and beyond?

Anant Adya: One broad statement is it is going to increase for sure. That's very clear from my perspective. And number two, the shift of cloud spending is going to happen more towards grow and disrupt, as I said in the first question, than on just save and optimize. So gone are those days when customers will say, "I want to get out of my data centers or retire my legacy technology infrastructure and just migrate the workloads to cloud." Right now the value of cloud is how do you launch new production solutions? How do you basically better the customer experience? How do you expand into new markets and focus on growing the revenues, right? So that's where the spend is happening.

Number two, the way it is going to change the future is AI, right? For every AI project, cloud is going to be the foundation. So like we say at Infosys, we first said customers have to think digital. So we said bring in a digital thinking mindset. And then we said with Cobalt, you have to establish a robust technology foundation, and that foundation is nothing but the cloud foundation. And as you now have the digital first mindset and you have established a solid cloud foundation, you need to start thinking of becoming AI first. So I think that is the future that we see. So I would say majority of your cloud consumption is going to be because of the AI projects, and that's the future of cloud.

Chad Watt: Anant, thank you very much for your time today.

Anant Adya: Thank you. Thank you for the opportunity.

Chad Watt: This interview is part of our collaboration with MIT Tech Review. In partnership with Infosys Cobalt. Visit our content hub at technologyreview.com to learn more about how businesses across the globe are moving from cloud chaos to cloud clarity. I'm Chad Watt with the Infosys Knowledge Institute. Until next time. Keep learning and keep sharing.