KonaWare empowers mobile workers

Execs discuss the market for mission-critical mobile apps

STARTUP KONAWARE, FOUNDED a year and a half ago, supplies a framework for deploying and managing applications that connect mobile workers to enterprise data. CEO Jim DiSanto, President Gary Portney, and Vice President of Services and Integration Faisal Mohamed met with InfoWorld Test Director Steve Gillmor, News Editor Mark Jones, and Editor at Large Ephraim Schwartz to talk about the need for delivering robust, mission-critical mobile apps.

InfoWorld: Tell us a bit about your vision of the mobile market.

021108hnkonaware_GP.gif
Portney: The way we see it, mobile and wireless have been completely over-promised and under-delivered over the last several years. We think there's a huge opportunity that's about to start around what we call "customer remote location applications," or mobile applications, and extending the office to all the workers that sit outside the four walls of the office. We see this space starting to take off over the next couple of years. Our [vision of the] evolution of the mobile marketplace is as follows: Generation One ran off proprietary networks that did one or two specific applications. [They were] typically homegrown and very services-oriented, [with] not a lot of off-the-shelf software that could address the problems. That is what FedEx did, what UPS did, what all these very big companies [did] where mobile was an occupational necessity. That technology is still here and companies are still spending the money around this proprietary technology. Generation Two was the promise of the wireless Web, where you had 2G networks. Entrepreneurs spent a lot of venture capital money to apply WAP to business problems. That should have never have been done. I know a lot of successful implementations of WAP in Europe, but they're all consumer-based and it all has to do with content. In the enterprise, it was pretty much dismissed.

DiSanto: What we're trying to get across with Generation Three is that there are some new technologies and new infrastructures that have come into being in the last six months or less, and this infrastructure development is what's going to provide the platform to launch enterprise mission-critical applications for mobile workers. In the past, we haven't had the infrastructure pieces in place, and because of that you couldn't get anything done. Still today this 2.5G stuff is fairly precarious.

InfoWorld: What do you see as being the critical pieces to the growth of this market?

021108hnkonaware_JD.gif
DiSanto: I think the biggest infrastructure issue is the IP wireless networks. Those networks need to be in place, they need to be reliable, they need to have the same coverage, and they need to work. [Another issue is] the mobile computing platforms. The hardware is becoming fairly solid and the price points have come way down -- Moore's Law is starting to work with the memory and the processor and the screens even. Sun is still working on the Java stuff, but .Net is going to be a formidable competitor in the mobile space. There's just no doubt about it, Microsoft is going to have their way there. For us, that's good news because it means there's something on a mobile device that's enterprise robust and it's something that an IT department can write software to without worrying about obsolescence the next day. Laptops, Tablet PCs, anything that moves outside the four walls of the enterprise -- these are the infrastructure pieces that need to be there.

InfoWorld: What needs to happen in the standards arena for all this to come together?

DiSanto: We see three operating systems and two programming standards which really matter [in mobile computing]. The two programming standards are going to be Java or .Net. Another key piece of this that hasn't really been explored is the adoption of the standard EAI connection mechanisms inside the enterprise. If you're going to build complex mobile applications that you distribute out to thousands of workers, you've got to be able to integrate these things through an EAI platform. The whole idea of building a native Palm or Pocket PC application that connects back to Siebel or connects back to SAP -- I can't believe companies keep trying to do that, including the owners of those applications, Siebel and SAP. [Users want] one mobile application that has access to all the back-end systems, so therefore this is not going to come from one of those big application providers. ... Our system is designed to drop into a J2EE or a Web services environment and to integrate through the EAI platform. What that means is that IT is going to be delivering all future services and applications through one of those standards; this helps them leverage that investment. They have already trained a ton of developers on J2EE: They already know how to program in Java, they already understand how to do XML. We're using those things; we don't want them to have to go to training schools.

InfoWorld: What is KonaWare bringing to the table?

021108hnkonaware_FM.gif
DiSanto: KonaWare is the only commercial software that we know of that can provide an XA-compliant transaction [XA is the OMG protocol for doing transactions across distributed systems] from the start of data entry, all the way back to the end system, regardless of what the network status is. Not just back to some intermediary server, but all the way back to SAP or PeopleSoft or Siebel. We guarantee the delivery of the transaction. A transaction that's IT robust, you can literally do a banking audit with it. An end-to-end transaction [is] much different than a guaranteed message. BlackBerry will guarantee you the e-mail will get through, but you can't do transactions with that because all the other attributes aren't there.

We have this very advanced administration system with our system. We're trying to be thought leaders in this area: How do you deploy and administer 5,000 little SAPs running around outside the firewall? If you deploy 5,000 SAP seats inside of a large enterprise and somebody has a problem, [they] call IT because it's inside the four walls. We have service workers deployed around the world -- they're in trucks, they're in remote offices, and they're out in the fields servicing transmission equipment [where] there's no IT staff to help them. We're talking about putting a transactional enterprise application like SAP, something sophisticated [that's] doing financials or order management or inventory or trouble ticketing and service management, on [mobile devices]. That's a whole new ballgame. And even more important [than helping] IT deliver a reliable service is [helping] keep their costs in control so they don't have 20 people running around solving problems all day, just escalating their cost. We have developed an algorithm, and part of our administration system is the ability to monitor how much money you're spending every day using these things. We give the IT department controls to trade off performance vs. cost. If I want more performance, I turn the dial one way; if I want to save money, I turn the dial the other way. When I expect more performance, I send more messages over the public networks, which cost money.

InfoWorld: What differentiates KonaWare from other companies trying to do this?

DiSanto: The first thing we do, starting at the top of the stack, we give you a smart client with an interesting user interface that makes mobile workers highly productive, allows them to work online and offline, and we call it Device Server Architecture, which basically provides the minimal amount of business logic and data on the client to get the work done while fully disconnected from the enterprise. And [you] also get the end-to-end, two-way transaction guarantee with our Kona shuttle and our Kona bridge. The shuttle resides on the device, the bridge resides back in the enterprise either in the DMZ or somewhere behind the firewall. These two pieces work together to provide the end-to-end two-way transaction guarantee. So if I want to send a transaction back into SAP from my Pocket PC, I can do that. And if I want to send it from SAP back out to the device, we can do that too. And we guarantee the transaction.

InfoWorld: If I am a mobile worker in company that uses SAP in the enterprise, do I have an SAP app for my handset or do I have to design an app to get the different kinds of information from the SAP?

DiSanto: We have this system called a Workbench which allows you to look at all the different functionality available on an SAP server and then map it over to the client. We call that "defining a view" and then you map it. The whole methodology we prescribe to is what we call "distilling and untethering the mobile apps." What you want to do is first distill it, figure out what it is that mobile worker needs. You've got to interview your mobile workers [and] figure out a flowchart on this application, what functionality you need. Then we're going to look at the SAP system and say, "What do we need to map over? We don't need the whole thing, we just need little pieces of it. We're not talking to the SAP system, we're talking to the EAI software that's connected into the SAP system. So the way SAP will do it is they'll expose their APIs through a Java server or maybe some Web services server. Those are the integration links we use. We don't ever go directly to the end-system [such as] Web Logic, WebSphere, Tibco, or Sun One [Open Net Environment].

InfoWorld: Are you relying on those systems to provide some of the business process management aspects of this?

DiSanto: It's not required, although it could be an additional benefit. You create a business process inside your EAI platform, because all those vendors are all selling that like crazy right now. And of course Microsoft is very strongly espousing that with .Net. All your business process should now be in the middle layer, literally.

Mohamed: If you have a set of business processes or business objects that you want to mobilize, it would be new processes. But if you know what objects you want to mobilize, you can easily bring them over to the device. So in Clarify, for example, if you have 20 pieces of information for an account, maybe the mobile worker only needs five pieces of information. You actually distill that and you bring the necessary information to the device. Same thing for SAP, same thing for PeopleSoft. They all provide you these objects that are exposed and then you can bring in whatever component that you need, a subset of that.

InfoWorld: How would a developer approach that layer of business processes?

DiSanto: With our product, anything that's written in Java, any Bean structure, can easily be introspected. We have a way of quickly producing a glue code between those Beans and our Kona bridge. It's [an] automatic generator.

Mohamed: We have an interface between the bridge and what we call our [JMS] server, which provides you a handle which allows you to completely integrate what we call a delegate. And it provides you all the information that you have from the device to the server to a delegate.

InfoWorld: Who are you targeting with your solution?

DiSanto: Our target [is] the early adapters; we call these MON customers, for Mobile Occupational Necessity. A lot of people complain enterprises aren't buying software right now. Well, you've got to go find those enterprises that are going to have to buy this stuff because there's a new infrastructure here. For example, there are 12 mail package and freight delivery companies in the Fortune 1000. UPS worked on this for the last three years and rolled out an application that's pretty nice now. They have tablet PCs from Fujitsu, they're wirelessly connected back to UPS. When a guy walks in and asks, "I want to send this package off to Spain, how much is that going to cost me if I need it there in two days?" [UPS] can find the answer for you right there. The FedEx guys can't do that -- they don't have information at their fingertips. Companies like FedEx have no choice but to adopt this stuff. There are 324 of these companies in these specialized verticals in the Fortune 1000 where we think they're going to be early adopters of this solution. We've contacted more than half of them -- we probably have 200 contacts now. We found close to 200,000 mobile users targeted for some sort of application in the next two years. In 2003 and 2004, 18 percent of the Fortune 1000 in the United States are going to adopt these solutions and roll them out. And then it goes up to two million users by early 2008 -- [in a] constrained economy.

InfoWorld: Are you betting on J2EE for the near term and waiting to see what you're going to do when .Net is available on the server?

DiSanto: That would be the right assessment. If you were going to build something like this over the last year and you're looking at a back-end integration strategy, where is Web services over the last year? It's forming, whereas J2EE has stuff out there [now]. I'm not going to say all of it's working, [but] we don't need it all to work. We just need the messaging components and the Beans, and that stuff is working; people have implemented it. It's there, so we're going to use it. Web services is the next thing we're working on, and the standards have literally just come out and implementations are scarce.

InfoWorld: How do you leverage the .Net platform?

Related:
1 2 Page 1
Page 1 of 2