Chris Doggett: Leaving Legacy Behind
Years of work on legacy policy administration systems led Chris Doggett to pioneer highly configurable, rules-based technology that has become the standard for the industry.
By the end of the 1990s, Chris Doggett had compiled 12 years of exeperience as an independent consultant working on decades-old legacy policy administration technology for companies such as AIG and ING. Before the new decade began, he resolved to build a new kind of policy administration system.
"I wanted to build a totally different system from the ground up, so in late 1999 I stopped consulting and worked full time on what became the AdminServer system," Doggett recalls. "Chris Prabhu [Doggett's technical partner] and I worked 18-hour days in the third-floor loft of his condo, and it took us about three years to come up with the first version of the system."
AdminServer set a new direction for life/annuity policy administration systems and influenced the course of other vendors across the insurance industry toward the development of highly configurable, "rules and tools" policy admin systems, according to Chad Hersh, a partner at New York-based research and advisory firm Novarica. "The solution's platform- and database-independence, scalability, and its embrace of XML showed carriers that mainframe, legacy systems weren't the only option," Hersh relates. "Another area that AdminServer helped bring into the mainstream was rapid product development capabilities, which along with rules and configurability are now table stakes for any successful policy administration system."
Doggett says his objective as he and Prabhu set out to build the system was to make it as flexible as possible in order to enable different products to run on the same set of code and even on the same server. Another requirement was that the system be a browser-based thin client in order to avoid the necessity to install software on client machines. "The system also needed to be non-state-specific, transaction-oriented and able to run transactions in real time at any time of day," Doggett adds. "It was to be written in the latest technology, which in 1999 meant VB6 and SQL Server."
A 'Beautiful' Discovery
Doggett recalls how his development team struggled with making rules sufficiently flexible to meet the objectives he and Prabhu had set. Then Duane Pressly, who was working for Microsoft at the time, stopped by Prabhu's loft one day and suggested that the team use XML and join an organization called ACORD -- the insurance industry's non-profit technology standards organization.
"When we added a free-form text blob to the business rules payload one day, I threw some XML in there and Chris was able to pick it up and have it drive a screen validation for numbers," Doggett recounts. "That simple validation was the event that set everything in motion. We came to the realization that we could add any amount of rules to a payload and that we had no practical limit to the rules that could go into a payload blob -- it was a beautiful thing."
Following several years of success in the insurance systems market, AdminServer was acquired by Oracle in 2008. The solution is now known as the Oracle Insurance Policy Administration System, which is a core component of Oracle's strategy to deliver a comprehensive, standards-based insurance enterprise software suite. "AdminServer's founders realized and acted on insurers' compelling need for a completely new type of policy administration system, one that could bring new levels of automation and flexibility across their enterprises and lines of business," comments Chuck Johnston, VP, strategy and alliances, Oracle Insurance.
Doggett -- who now spends his time running a resurrected speakeasy in Philadelphia, the Franklin Mortgage Investment Company -- attributes the successful development of the innovative AdminServer system to five factors: the smallness of the development team, single-minded focus, the choice of an open technology environment, the lack of any "oversight and governance committee," and, lastly, how poorly the work paid -- "This was actually a good thing," he insists. "We needed to get it done and sell it to someone or we were going to have to look for real jobs."