By Charles King, Pund-IT, Inc. April 15, 2015
I first met Dr. Jai Menon in the late 1990s at IBM’s Almaden Research Lab where he led the storage team. Dr. Menon was something of a legend around IBM, having joined the company in 1982 where he had been instrumental in the development of RAID (Redundant Array of Independent Disks) technologies. Dr. Menon has written 23 RAID papers and holds 27 RAID patents, work that led to him receiving the 2002 IEEE W. Wallace McDowell Award and 2006 IEEE Reynold B. Johnson Information Storage Systems Award.
Given his achievements, it was no surprise when in 2001 Dr. Menon was named an IBM Fellow, the highest engineering accolade in a company known for its commitment to research and development. Other achievements followed, including being named VP and CTO in IBM’s Systems and Technology Group. Many IBM Fellows are content to finish out their careers with the company so it was surprising in 2012 to hear that Dr. Menon had joined Dell as the company’s first chief research officer and leader of the new Dell Research organization.
In that role, Dr. Menon is responsible for driving organic, long-range, disruptive innovation bridging all of Dell’s focus areas to inform the company’s technology strategy and enhance its solution portfolio. Recently, I had the opportunity to interview Dr. Menon about his work at Dell, the company’s approach to research and its commitment to technological innovation.
CK: Before joining Dell, you were at IBM for 30 years as a CTO, VP of Technical Strategy, and also as an IBM Fellow. I’m curious about why you joined Dell, and how your current role differs from previous positions at IBM Research?
JM: What attracted me to Dell were a couple of things. First, it was an opportunity to innovate in a different kind of way. Dell works with 10 million small and medium businesses with a focus on democratizing technology and making it available to more people and organizations. That requires a different kind of innovation, and I found that challenge was really attractive to me.
CK: How about an example?
JM: A classic example is something we announced very recently, the Dell Storage SC4020 all-in-one array. At $25,000, it’s the lowest priced entry-level/midrange, all flash array in the industry from a major vendor. But price isn’t everything. It offers capabilities that might require a very expensive box if you were buying it from another vendor. It has the highest performance of any array under $100,000 and we achieved that with innovations including mixing lower cost flash with small amounts of higher performing flash, and migrating data in a very clever way to make it feasible to provide that level of performance and capability at a very attractive price point.
CK: Was Dell’s leadership also a factor?
JM: I have to be honest and say that one of the first people I interviewed with at Dell was Michael himself, and his passion for technology and what it can do for society was extremely inspiring to me. It was very different from the kinds of conversations I’ve had with other CEO’s because he’s so knowledgeable about technology. For a technologist like me, working for a CEO that can have a technology-oriented conversation about storage or networking or security was very compelling. I mean, just take any aspect of IT and Michael is very aware of what is going on, what vendors are doing, what is current, what is coming. It was powerful to sit in a room with a CEO and have that kind of conversation; it was so different from discussions I’ve had with other CEOs.
CK: That’s an interesting point, because often, particularly in the start-up world, a company will be initially driven by technologically oriented founders and then at some point they’ll decide it’s time to bring in “the grownups”, the people who are so-called business experts. You can understand that from an investment standpoint, because the business people are much more comfortable in speaking the language of that constituency. But often there seems to be a dividing line between the business and innovation that doesn’t usually work in the best interest of innovation.
JM: Exactly. Michael is a true business leader but one who has never lost his passion for technology and how it can make a difference. He is able to hold his own with technologists. That was very compelling reason for me to join Dell.
CK: Going back to the democratization of technology, what part does commodity technology play in that? Say, rather than building a brand new widget, putting together existing commodity widgets in new or innovative ways?
JM: Every time I’ve thought about this problem I’ve determined that it’s not really about commodity technology. When you want to democratize something, it can require significant innovation. Let’s take big data as an example. If you think about big data for large customers, you’ll go down a particular mindset because they have the ability to hire large numbers of expensive data scientists. They also have lots of data, and the more data you have, the better job you can do with analysis. But if you think in terms of making big data available to the 10 million small and medium businesses that Dell interacts with, you have to remember that many of them cannot afford data scientists and they’re also disadvantaged in terms of how much data they might have relative to their bigger peers. So when we think about how to democratize big data for that group, we have to first think about how to make it work. How do I make a solution so easy to use that you don’t necessarily need a data scientist in the picture? It’s a different kind of innovation. Similarly, how might I get a number of small businesses to agglomerate their data in an anonymized way so that you can provide useful information back to any one of the SMBs contributing to the bigger pool.
CK: Sounds like a fundamentally different exercise.
JM: It’s a different set of things that your mind has to go through, but it’s still innovation. It’s just that it makes you think differently than you would if your sole purpose was to build let’s say the biggest thing that only the top three customers in the world might be interested in. Both approaches are innovative, both are exciting, both are challenging. At Dell I have the opportunity to think about both SMBs and large enterprise customers. I do this within our broader Innovation team at Dell, which focuses on a practical innovation lifecycle in four essential investment areas: entrepreneurship (human capital), strategic alliances (partnership capital), M&A and venture capital (financial capital), and R&D (technology capital), where my team sits.
CK: Interesting. Back to your comment before, is the pooling of anonymized data something that Dell is actively involved in? Would the data be pooled on an industry by industry basis, or regionally?
JM: We’re looking at different options and thinking about the technology that would securely enable anonymized, pooled data. Clearly we have some verticals where Dell has a lot of strength, like healthcare. The DCCA, Dell’s Clinical Cloud Archive, is the largest medical image database in the world. That’s an industry vertical example, but you can see many models. I think that some of the technical questions around how you anonymize it and how do you handle that in a secure way are the same in every case.
CK: Interesting. You know, you mentioned Dell’s history of innovation in servers. Are there other areas related to that you’d like to discuss or other efforts where Dell has done interesting and innovative work?
JM: Staying on servers for a minute, Dell recognizing the trend toward scale out and hyperscale computing made us one of the earliest – if not the earliest – vendor to create a whole line of offerings to go after that opportunity with our Data Center Solutions (DCS) division. That’s an example of recognizing where the puck was going and then becoming a leader in that space with significant market share. One of the things I remember from my earliest interviews at Dell was really being blown away by some of the modular data center technologies that Dell had and the PUE (power usage effectiveness) of 1.03 that Dell was accomplishing.
CK: Why was that so impressive?
JM: Well, you hear about Google or Yahoo achieving PUE (power usage effectiveness) ratings of 1.07, which is impressive. Then I came to Dell and saw that they were beating what those leading companies were doing. I found that very innovative.
CK: How about other areas?
JM: The work that Dell did in terms of iSCSI and being a leader and now being a major owner of that space was very innovative. Timing, by the way is very important in being innovative. You can’t be too early or too late. You can be wrong on both counts and so being just right is important. I think Dell’s timing was just right with iSCSI. That’s an example of something I had recognized as an opportunity before I came to Dell. When I saw that opportunity years back, it was too early, before the right standards were in place, before the customers were ready to accept this kind of change that was coming. Dell not only identified the direction correctly, but really timed it well so that now they are a leader there.
CK: How about networking? That’s an area where a lot of vendors are jostling for position.
JM: Dell is a leader in open networking – a disruptive approach in which we allow customers a choice of network operating systems to run on our switches. The network operating system can be from Dell or it can be from one of our partners in our open networking ecosystem, such as Cumulus or Big Switch. This allows our customers to tailor their networks for their application needs in a way that was not available to them before. I would also say that the leaf and spine architectures (a network topology in which a series of leaf switches form the access layer and are fully meshed to a series of spine switches) that are really critical to large scale out data centers is another great example of innovation. Dell was an early leader in going after that networking architecture and leveraging its full value.
CK: That actually leads me into the next question, which has to do with the relative benefits of organic internal development efforts and external mergers and acquisitions. Dell’s iSCSI and networking efforts have been informed by pretty substantial and innovative acquisitions. Do you think that one approach is better than the other, or can they be complementary?
JM: They’re very complementary to one another. Organic approaches make a lot of sense when they’re related to an existing company strength. DCS is a good example of that; Dell’s strength in servers enabled it to develop new servers with a different set of attributes. But if you’re not in, say, the networking business at all and you’re trying to enter into that market, then making a Force 10 type of acquisition makes a lot of sense. Following that acquisition, the open networking initiative that followed was an organic Dell innovation.
CK: So it’s not always an “either/or” decision?
JM: I don’t know that I can cleanly say that it’s always organic for this situation and inorganic for that situation, but Dell Research partners very closely with the M&A and Ventures teams, so we are always passing data back and forth. They might see something interesting and they might ask us to do an evaluation for them, or vice versa. In some areas where we’re looking many years ahead and see something really interesting, we might work with an external start-up for six months or a year to evaluate an opportunity and decide whether it looks really good.
CK: It sounds like a dynamic relationship.
JM: That kind of back and forth is very important to us. It also helps us evaluate areas in which there is significant external innovation and maybe it doesn’t make sense to do things organically, versus those areas where it makes sense for us to pursue. So it helps us to shape what we do.
CK: What part, if any, do partners play in this? Is there room for doing research and development with partners?
JM: Very much so. I think that it’s critical to what we do in our research group. For example, Dell does not build microprocessors, so Intel is an amazingly important partner. We need to know where our partners are going and work closely with them to help shape and influence their direction. Knowing how a partner like Intel is thinking about next generation processors or next generation memory technologies is very important to us.
CK: Don’t Dell’s competitors follow a similar path?
JM: This is another advantage of working at a company like Dell, because in some areas we have the freedom to explore, work with our partners, understand their roadmaps, see where everybody’s going, influence all of that but also be able to pick and choose the best of what is out there. One of the things that I found in other companies that make their own investments in chips and memory technologies is that they are often constrained to choose the technology they have invested in to solve particular problems. So, if you’ve invested a lot in a particular kind of memory technology like the memristor, then, you know you want the memristor to be the answer. Whereas we have the liberty and luxury of being able to look at different memory technologies like, say, phase change memory, and weigh it objectively, and then decide which is better in an unconstrained way. So I’m not constrained by having to force myself into deciding X or Y or Z is the winner, just because I have invested heavily in X, Y or Z. Instead, we can base our design decisions and innovations on what’s best for our customers.
CK: Well, if you buy or build an elephant gun it’s inevitable that at some point you have to go elephant hunting.
JM: Exactly. But that also helps us focus our work and leverage the work of our partners. Clearly the work that we do with our partners is very important, and not just for Dell Research but for Dell in general. Democratizing the cloud requires us to have partnerships with cloud vendors – with Amazon and Google and so forth. Our Dell Cloud Marketplace solution is another way of democratizing the cloud.
CK: How so?
JM: Cloud is a complicated thing. If you are a small business, do you really want to have to figure out whether your work load runs better on Amazon or Google? Or do you want to come to Dell’s portal which has already tested and determined cloud performance? When you think about making cloud scalable to many, many, many businesses of all sizes, that requires partnerships.
CK: It’s an interesting business model for Dell to act as an adjudicator of cloud providers, especially since in the last year or so many large traditional systems vendors have gotten into the cloud business themselves. Whereas by staying out of the cloud business, Dell can act more objectively.
JM: Right. Exactly. We support VMware, Microsoft and Red Hat for private clouds on our servers, and are also objective about public cloud providers.
CK: Google is famous for supporting what they call moon shot projects that aim to balance high risk with potentially large rewards. Are you and Dell pursuing similar kinds of projects? Related to that, what the importance is of balancing technical innovation against commercial opportunity for Dell?
JM: Let me approach this from a slightly different direction. Dell Research has developed what we are calling the Dell Technology Outlook (DTO), which discusses where we see things going and the disruptions we foresee in the coming five years. We also pursue research projects that support the DTO vision. Specifically, we’re looking at four areas or pillars which align very nicely with the way that our customers are thinking: Transform, Connect, Inform and Protect. Those are the areas that we’re focused on.
CK: How does that play out?
JM: Transform focuses on where the data center is going. Connect revolves around mobility and the Internet of Things (IoT). Inform centers on big data, and Protect focuses on what’s happening in the security space. We’ve developed fairly clear points of view on where we see the world going in each of those four areas and we are aligning our research projects to make sure Dell and our customers are prepared for the disruptions we predict in each of the four pillars. The Outlook is very much about the implications of these issues for our customers. And when we conduct research, we are always looking for areas where we can develop a proof of concept (POC), where the customer has a current pain point and they want to work with us to solve it.
CK: That seems very practical, that the focus is very much on replicable, commercial solutions.
JM: I’m not looking for everything that we work on to be successful or create a new product or a new business in Dell. I want our success rate to be somewhere between 50-70 percent. I don’t want a 100 percent success rate, because then I’m not taking any risks. I’m not also trying to have a whole bunch of moon shots where, say, only 1 percent will succeed, but if it succeeds it might make a billion dollar company. Our approach is finding a happy medium, and I think that’s the right balance for where we are as a company.
CK: I’d like to change tacks. I’m curious about how important patents and proprietary technologies are in the process of R&D generating innovation. I’m also curious as to whether you think the number of patents that a company creates annually is an adequate or accurate measure of its potential to be innovative?
JM: Patents are important, particularly from a freedom of action perspective. But I don’t believe that the sheer number of patents that a company generates is an indication of their innovation capability.
CK: Why so?
JM: Because a lot of companies generate many patents, but 90 percent of them have never seen the light of day in a practical product. To me, innovation is the intersection of the invention (which is what the patent is) with its application to and usage in making a difference to society.
CK: In the real world.
JM: Yes, in the real world. If an invention just exists as a piece of paper and never has an influence on the real world, then it’s not innovation to me. So invention is a part of innovation, you can’t have innovation without it. But true innovation is the practical application of invention in to the real world.
CK: So innovation is almost a measure of the impact of an idea after it leaves the lab and escapes into the wild.
JM: The impact of an idea is very important. I tell my team that we should measure ourselves by what we create. Did we improve a product? Did we create a new product for Dell? Did we create a new business for Dell? Each step is increasingly harder.
CK: Can you elaborate on that?
JM: You can take an existing product and make it a little better, bring in some new features and take it to market faster. Or you can create a whole new product, but in an area that we already know. Or, you create a whole new business in a new area, and that’s even harder.
CK: Can you cite any examples?
JM: In that last category – creating a new business – I would point to some of the work we’ve done to date with respect to new products targeted at telecom companies in an area called network function virtualization (NFV), and something that within Dell Research we’re calling the high velocity cloud. We saw that as a business opportunity, and we did early leadership technical work in Dell Research. In October of last year, Dell created a new business organization to go after this opportunity based on our efforts. I don’t expect to do that all the time or even very often, but that’s the highest level of impact that our work can have. I’m very proud of the fact that we’ve been able to achieve this so soon after formalizing our organization. But you know, you can also have important external influence in other ways.
CK: Like what?
JM: Leaving a mark on the industry in some important way, such as a standard that everybody uses and adopts that could drive an entire industry. So it’s more than just one product or one business. Ultimately, in all these metrics, simply writing a paper is not good enough. Writing a patent is not good enough. Not by my standards. You’ve got to translate that into something that has value in the world around you.
CK: Did Dell Research exist as an entity before you joined Dell?
JM: Innovation has always been alive and well at Dell. The company has innovated to make technology more accessible to a wider range of organizations, simpler to adopt and operate, and more affordable and easier to own. Innovation has always happened within each of the business units and it continues to happen there. Dell Research was formally created in 2013 to augment this effort that was happening within the business units. We focus on delivering long-range, disruptive, pan-company research and development to inform and influence Dell’s technology strategy as far as 10 years out from now.
CK: Was there a particular factor that motivated Dell to form a dedicated Research division?
JM: I think that there were a couple of motivating factors. First, as Dell went private, it made even more sense to start Dell Research, because we had a bit more freedom to invest in what we thought were the right things for customers without always having to justify them on a quarterly basis. So the timing of Dell Research matched very nicely with Dell as a private company. Secondly, Dell has also been in this transition towards being an end-to-end solutions provider, and that requires an entity that looks across all of the business units. So, as I mentioned, there is a lot of innovation going on within each individual business unit. But if I’m trying to do things end-to-end, whether it’s in security or mobility or cloud, where we’re piecing together hardware and software and services, you have to have a pan-Dell lens. Dell Research is an organization that is truly pan-Dell in its thinking. We can bring together those different pieces and innovate through making connections.
CK: I’m curious about how Dell Research is organized. Is your team centered here in Silicon Valley, or if you have facilities elsewhere? I’m also curious about how large the group is, from the standpoint of personnel.
JM: We’re organized along the four themes that I talked about. So we have a group that’s thinking about what the next generation data center will look like, one focused on security issues, one around big data, and one looking at mobility and IoT. Location-wise, we have people here in Silicon Valley as well as employees at our headquarters in Round Rock. We’re also expanding internationally. Last fall, I visited India, where Dell has a large R&D group, and so we’ve carved out a Dell Research team there.
CK: Are you working with universities as well?
JM: Yes, very much so. Most of our work is with U.S. universities at the moment, but there are also very good institutions in China and India. As an example, in the U.S., we’re working mainly with a networking group at the University of California at Berkeley, led by Prof. Scott Shenker, who is considered by many to be the father of software-defined networking (SDN). He recently visited us and we had an exchange of research ideas. We also have a professor who has taken a sabbatical from the University of Southern California working with us, marking our first sabbatical to Dell Research. So the word is already getting out that we are an interesting place to partner and work with. We are also working with some universities in Texas on next generation memory technology and we have partnered with Clemson University for a $10 million dollar NSF grant for work on something called Cloud Labs.
CK: Your work with universities sounds like a dynamic part of the program.
JM: I view it as an important complement to the organic work that we do here. There is a lot of talent in universities. As long as we can align the research to the areas that are strategic to Dell, there is a lot of value that can be gained by these kind of interactions. So it’s very important that we continue that.
CK: Are there any restrictions on the work or projects that you pursue at Dell Research?
JM: No, as long as projects are associated with our four themes of Transform, Connect, Inform and Protect, there are no other ground rules. Like I said, we have some metrics for success so the researchers keep those goals in mind. At the end of the day it’s about making an impact on Dell and having influence on society in some way. There are so many ways to have influence. We set a world record last year with our telco infrastructure research project. That’s a very worthy goal for us. So there are no other kinds of restrictions as long as our research aligns with our business priorities.
CK: Can you share some background on any of your current projects?
JM: Sure, I’ll hit on the highlights in our four thematic areas so you have a sense of what we’re doing. In IT infrastructure, under our Transform pillar, we are focused on what we’re calling the software-based data center.
CK: How is that different from a software-defined data center?
JM: In the software-defined data center, software is used to orchestrate, manage and provision resources in a very agile way. That makes perfect sense and it’s very important, but the resources themselves could be custom hardware resources that happen to have appropriate API’s that allows the custom hardware to be managed and manipulated. In a software-based data center, we’re still using software to manage all of the resources and orchestrate them, but the resources themselves are largely software running on servers. That is, you don’t really need custom hardware for a lot of things that once required it.
CK: Can you give me an example?
JM: Sure. The telco world record that I just mentioned is a good one. To support a million mobile phone users in a city like San Jose, telecoms today use highly customized boxes that cost about a quarter of a million dollars. We proved that you can support a million mobile phone users all doing mobile and video traffic – with a quarter rack of standard Dell servers, which is a lot cheaper than buying specialized boxes for a quarter of a million dollars.
CK: That’s pretty impressive.
JM: We see the software-based data center as part of a trend away from depending on special purpose, customized boxes. Commodity servers are getting fast enough to do a lot of things that they weren’t able to do few years ago. Moore’s Law is still pretty much alive and well, at least through the 2020 time frame as we see it. So that’s one of the directions we’re going. We’re also looking a lot at next generation memory technologies, for example, what comes after flash, how soon will it be here, and how will it impact our customers.
CK: How about security?
JM: In the security space, we’re looking at predictive kinds of technologies. I like to say that security today is a little bit like changing the lock on your house after somebody burgles it. But you want to move to a model where you become more proactive and change your lock when your neighbor’s house gets burgled. We’re starting to do that with our SonicWall firewall appliances. We have two million of them around the world and if we see an attack going on in China or somewhere else, we could update other unaffected firewall appliances immediately. But ultimately you want to get even more predictive that that. You want to be able to recognize and respond as an attack starts and be able to prevent it. Essentially, you’re making the lock a better lock even as you hear the burglar trying to pick the lock.
CK: Mobile security seems like a very hot area just now.
JM: Yes, it is. The Global Technology Adoption Index (GTAI) that Dell recently commissioned found that around half of our customers are concerned about security on mobile devices, particularly having them fall into the wrong hands. What we’re doing to address that concern is what we call “continuous authentication,” meaning that not only do you authenticate yourself once when you log in or type in your password, but then the device continuously authenticates you based on your pattern of use, such as the way you would swipe and the pressure that you apply on the touch display. The way you would do this is different than the way that I would do it. So if a device is trained for me, it will simply lock out you or anyone else who tries to use it.
CK: That sounds almost like a forensic approach to personalized security
JM: Yes, it is a forensic approach.
CK: How about other mobile projects?
JM: In the mobility area, one thing we’re looking at is BYOD. A lot of people have been highly focused on mobility management for BYOD, but the work has nearly all been related to security. That’s important, but what about ease of use mixed with that security? So what we’re doing here is a project we call seamless mobility. The idea is that as you walk around a Wi-Fi enabled environment – a business campus, a hospital, a university campus – sometimes you’ll have good Wi-Fi and at other times you might have good cellular signal, and sometimes both may be available to you. The problem is that if you’re using a mobile app with a cellular connection and you move around, when cellular drops, your application drops with it. With our approach, if cellular drops but a Wi-Fi signal is available, the application continues to run because in the background, we will switch you from cellular to Wi-Fi. Seamless mobility is always measuring the relative strengths of Wi-Fi and cellular signals, and not just strength but latency, too, which is important.
CK: As anyone using public Wi-Fi knows.
JM: Yes. You can be in a Starbucks and it’ll tell you strength is fantastic with five bars. But there’s another hundred people there all accessing things, so your latency is actually not very good. Then the five bars don’t mean anything.
JM: So that’s why seamless mobility measures latency, as well. And you can actually set it up so that you can decide what policy you want to apply to any specific application. You might decide that you really want it to be seamless just for email, and it will just seamlessly move you back and forth from Wi-Fi to cellular for that application. There may be other applications where what you really want is aggregation – this would be for applications that need so much bandwidth that you want to consume everything you can get, so if both Wi-Fi and cellular are available, you want to use both. You can also relegate certain applications to just one or the other. If your data plan is very expensive, you can offload to Wi-Fi whenever you can and avoid your using up your cellular GBs.
CK: Or, from a business perspective, you might have an application that should only be shared internally. So you could dedicate access to that to the Wi-Fi alone and make it inaccessible via cellular.
JM: Exactly. Our vision in the Dell Technology Outlook is that BYOD will evolve to be not just about security but about security combined with ease of use. Seamless mobility is an example of that ease of use focus that is also very secure. We do encryption end-to-end, so even if the Wi-Fi network is unsafe, our end to end encryption makes interactions consistently secure whether you’re at an internet cafe or you’re inside the firewall, or you’re at home.
CK: In a way, mobility touches most of the other focus areas.
JM: It does. Another mobility and user interface design area that we’re interested in is pushing the envelope in terms of understanding the user that’s in front of the device.
CK: Why so?
JM: Because the more you know about the user, the better job you’ll be able to do for that user. And one of the next dimensions of knowing more about the user is about knowing more about his or her emotional state, including demographic elements, attentiveness and so on. This could be useful in a number of ways. For example, if you’re driving a car or on an assembly line, the fact that you are no longer attentive could be important to know in terms of safety or productivity. If you’re playing a game and we can sense that you’re bored, we could ratchet up the challenge level of the game and make it less boring and more entertaining. If you’re in a work setting and are in deep concentration, distractions like phone calls could be set to go directly to voicemail so they don’t interrupt you.
JM: From a demographics point of view, if we can tell that a child is using the computer, then we could shut off access to certain websites noted by parents. If we can tell that the user is an elderly person, we might automatically increase the font size on the screen. There are a lot of things that you can do with this kind of technology, including emotion sensing, so that’s been the intent and the driver for why we’re starting to look at these kind of things.
CK: This seems like a variation on the security lockout that you demonstrated to me earlier. A finer grained form of personalization.
JM: You might call it that. One research area we’re exploring with partners uses facial recognition to determine your mood. We also have voice tone based input to help analyze the stress levels in your voice and be able to, say, call 911 if you sound panicked. We actually used the facial recognition mood sensing technology at a TEDx conference that Dell sponsored in Amsterdam last November, where we had a number of cameras tracking people in the front rows, looking at their reaction to the speakers. That way, we could give good feedback to the speaker about how well they were doing, how well they were going over, and also how different demographics were responding. For example, you could be doing really well with women, but you’re really bombing with the men. Or, you’re doing well with the teenagers but not with the older people in the audience.
CK: That seems like it could have numerous commercial applications.
JM: Sure. You could see extrapolating that to teachers in particular settings, looking at their pedagogical style to be able to say this works, this doesn’t work. Every time I give a talk it’d be good to know how I’m doing. There are a number of possibilities here.
CK: What about analytics?
JM: On the analytics side, we are looking at predictive and prescriptive analytics, but we are more focused on the prescriptive side. In particular, we’re looking at health care as a vertical to target applications. Our StatSoft acquisition included a tool called Statistica that works well for predictive healthcare analytics. So we’re trying to enhance those assets with work in Dell Research to deliver better readmissions modeling; the risk of somebody getting re-admitted if they were to leave the hospital today. I mentioned earlier that Dell has the world’s largest medical image database, so looking at what we can do in terms of helping radiologists discover certain conditions, along with impacting general population health. For example, why is this community doing better than that community? Why are there more suicides here versus over there, and is it related to the number of social workers in the community or other factors?
CK: This all sounds extremely fascinating. I look forward to hearing more about your work in the coming years and how Dell Research continues to make a positive impact.