Armed Forces Insurance CIO Talks Data

Drew Mazeitis has been CIO and VP of marketing for Armed Forces Insurance (AFI, Leavenworth, Kan.) since 2008. He talked to Insurance & Technology about three projects he has in the works: a mainframe migration, construction of a data warehouse and a developing cloud strategy.

Insurance & Technology: You're in the middle of extracting your data from AFI's mainframe and rehosting it on a Windows Server platform. Can you talk about why AFI decided in 1997 to go from a mainframe-based system to a Windows Server-based system?

Drew Mazeitis: The mainframe costs AFI somewhere between $800,000 and $1 million just to run. I've decided to take our mainframe applications and move them to a Windows Server platform as a hedge while implement a new policy administration system. The development for it is complete for all of our homeowner applications, and is in testing right now.

I&T: How far along are you?

DM: The entire [rehosting] project is scheduled to be completed in May. If I can reduce my mainframe maintenance costs by $600,000 per year, I get that benefit every year that I'm waiting. Of course, ironically, the sooner we get the new policy system up, the less of an ROI I get.

I&T: But even then, it gets a lot of the work done in advance that's required to implement a new policy system, correct?

DM: That's right. Not only do I get all of my data off the mainframe and turn it off to save my lease and maintenance costs, I also get all my data on SQL server, which is where all my other data is. I get my mainframe programmers programming in visual studio, which is where all my other programmers are. They can take advantage of the efficiencies, the debug tools, the source control. I get everybody to a common database platform, development platform. I get to move all my data to a common SAN (storage area network; AFI implemented one from NetApp last year.)

I&T: You're also working on implementing a data warehouse and business intelligence platform. Are these initiatives related?

DM: You would've thought that was true. We started the rehosting project shortly after the data project, and I assumed that we could use the same extracts for each. It turns out that the level of detail I need for the data warehouse isn't as much as I need to get the applications to work in the rehosting. It's every field, every record. For data warehouse I just pull a transaction. It's like 70% of the same stuff.

I&T: What did you do to work around that?

DM: We used the warehouse data when we could, and when we needed more we created more tables for the mainframe data. It was completed December 31.

I&T: What are you looking to do by building this capability?

DM: We want to understand our business better and understand our customers better, and we also have reporting requirements. Our data warehouse project was started as a way to meet all those needs. We also wanted to consolidate reporting across our org so we have common numbers and terms. For example, what does premium mean: is it direct, total, written? There are a lot of different ways to calculate some of the different numbers in the organization. For example, a new member: marketing has a definition for when someone is added that's different from, say, our billing and financial dept. Is it when they sign their application or get the first bill? This gets us to common methodology.

I&T: Have you been able to get some of the insights that you were after?

DM: A lot of our decision making was done based on intuition and experience, and not enough was based on data. The bulk of the project was to extract data from many sources: our mainframe, phone system, billing system and website so we could compare them to each other in ways they couldn't before. It's based in SQL, using SharePoint and PerformancePoint as the front-end for some of the dashboarding. The business intelligence helps us pull together our data that allows us to ask of the data: To whom should we be marketing? What's our target customer? What geographies have been successful in the past? It gives us a framework to extract the answers to questions you want to ask.

I&T: You've been exploring cloud computing for some of your e-mail and HR needs, using Microsoft BPOS. Why do you think insurers are getting more comfortable using cloud services?

DM: People understand now the exchange that comes in with the cloud. As a mid- to small-size carrier, there are things that because of my scale make the cloud more attractive to me than, say, someone the size of an AIG. They have the resources to build data centers that have two generators and three Internet connections coming it. If I can find someone that meets our needs in terms of security and scalability, I'm going to go with it.

I&T: Has it worked out like you hoped?

DM: Yes. Now, we don't need to keep an Exchange admin on staff, and we don't have some of the headaches associated with managing e-mail. We also get the benefit of redundancy in terms of having our e-mails hosted is something that's more tight and bulletproof than something we could build ourselves.

I&T: That's a lot of change in the time you've been there.

DM: When I started two years ago, there was a lot of opportunity to refresh and update things. To me, the Internet is the great equalizer — it's where small companies can compete with the big guys. You put good tools on the site and it can be a differentiator. We just launched a new website on December 27, and it's going to be a platform we're going to continue to build on. It's something that we just started probably 18 months ago.

Nathan Golia is senior editor of Insurance & Technology. He joined the publication in 2010 as associate editor and covers all aspects of the nexus between insurance and information technology, including mobility, distribution, core systems, customer interaction, and risk ... View Full Bio