The Social Security Administration is shedding its reputation as an agency stuck in the 1980s when it comes to IT.
SSA cobbled together a few leftover computers and took advantage of an open source platform to develop a new approach to conquering its big data problem and address a huge improper payment challenge.
“The project was really innovating the audit and integrity applications and using big data to do that,” said Ron Sikes, SSA’s director of business intelligence and analytics. “This was innovative for us because it’s using non-standard technology. We usually deal with mainframes, relationship databases and things like that. We had to go completely outside our normal scope to be able to process billions and billions of records very quickly.”
SSA started with one excess desktop computer to test out the approach using Hadoop — an open source platform that relies on distributed storage and distributed processing — and a custom enterprise Web service that establishes connectivity between Hadoop and the WebSphere graphical user interface.
Sikes said SSA soon expanded the number of PCs to six, and then to 26.
“We were processing 60 times faster than our multi-million dollar mainframes, which is taking hours to minutes,” Sikes said.
Ann Amrhein, SSA’s deputy associate commissioner for the Office of Earning, Enumeration and Administrative Systems, said the end result of this effort was not only to stop improper payments, but take advantage of the data to make better decisions.
“This opened up a whole world of analytics and big data to SSA,” she said. “This was a way for us to begin to explore new architecture to be able to find anomalies, to predict situations much more quickly than we were able to do in the past..”
In fiscal 2013, SSA paid out about $6.8 billion in improper payments, mostly from its supplemental security income. The Office of Management and Budget says on its PaymentAccuracy.gov portal that the SSI program has an improper payment rate of 8.1 percent — the sixth highest among the top 13 high-error rate programs.
The New York City Department of Transportation’s NYCStreets Permit Management System won the grand prize of the day for its effort to replace a 30-year-old mainframe system with modernized self-service approach.
The system processes about 400,000 street excavation permits a year, and now contractors can apply online, receive approved permits via pdf format for printing, and DoT employees can evaluate conflicts using a Web map interface that uses Google Street View.
In all, 30 programs received recognition as finalists out of more than 100 entrants with the federal community voting on the eight to present at the awards event and picking the grand prize winner.
For SSA, earning a top eight nomination isn’t just a recognition of a successful use of technology for one program, but a potential springboard toward a broader innovation culture.
“It’s a step in the right direction. One of the things SSA has done quite well for many years in protecting its data,” Amrhein said. “This gives us the tools that we need moving forward to be much more aggressive in that protection, and it opens up a whole new world for us in addressing more than just fraud and abuse, but looking at ways that we can look at service to the customer in different ways than we’ve done today.”
She said SSA needs to find ways to interact with its customers more efficiently and effectively, and in a way they expect.
Sikes added the new technology approach also helps ensure the data security as SSA modernizes its customer interactions.
“It started out as a very small team, a very small budget and very dedicated people. They worked very hard on this. They worked late nights, weekends to try to get this done,” he said. “As we get recognized for this, it expands throughout the agency and probably across agencies, we will be able to help them as well. It really starts fostering that sense of innovation is good and we can do things that we haven’t done before.”
Sikes said SSA has had initial discussions about the technology with the Centers for Medicare and Medicaid Services and with some members of the intelligence community.
Amrhein said SSA is examining other potential areas to apply this technology approach to in the coming months.
“The long term vision is to increase the usage of this to be able to do predictive analytics, to be able to store all the data in one place so it’s easier to access for everyone inside the SSA community,” Sikes said. “We also want to be able to start pushing advanced analytics, looking at different algorithms that we can do very quickly and really just increase the capacity of all the people in SSA to look at data quickly along with preventing fraud.”