Better procurement can feed transition to 21st-century federal IT, Kundra says
Federal CIO Vivek Kundra talks about administration's new priorities
Since becoming the first official federal chief information officer in March, Vivek Kundra has not dallied. Following President Barack Obama's call for transparent government, Kundra, former chief technology officer for Washington, D.C., quickly set up Data.gov, a clearinghouse of raw government data feeds. And to follow up on Obama’s calls for federal procurement reform, Kundra is now looking at ways to simplify and speed the federal information technology procurement process. In July, at an event hosted by the National Defense University, he said the General Services Administration would establish a storefront that other agencies could use to rapidly tap into cloud computing services
GCN: The George W. Bush administration launched a number of cross-agency lines-of-business projects for common agency duties, such as financial, human resources grants and case management. Where do these initiatives stand with the new administration?
Vivek Kundra: I think there is an important commitment to [the federal lines of business], but you have to look at the entire architecture. We need to move to an environment where we are making very common-sense decisions. If you live in a household with a family, you are not going to have different solutions for a problem you can solve commonly. You are not going to go out and build separate infrastructures.
University of Maryland confirms Kundra's credentials
Future online storefront to let agencies buy computing services
Kundra talks Data.gov
Kundra courts the risk of innovation
In the government, what we need to start doing is first look at what is happening at the departmental level and then scale that up to an architectural model at the federal level — the [Financial Management] Line of Business, the [IT] Infrastructure Line of Business. We need to think through what we need to do in terms of infrastructure. What is the agency's mission? How does it tie back to the federal enterprise architecture data models?
I think we also need to ask different questions. Let's take the infrastructure line of business. Why not look at some of these game-changing technologies, like cloud computing? Do we really need to spend billions of dollars in data centers across the federal government? Do we really need to use up all this energy when we can do it in a lighter-weight way. What are some of the opportunities out there?
In that context, what is happening with the lines of business is that there is a lot of debate around thinking through some of the newer technologies that we could leverage. What about a migration into a shared-services model?
I also recognize that on the back end, there are some really complicated legacy investments, coupled with business processes that are fairly old. The harder part of this is to make sure the business processes are re-engineered. How can you continue to run on businesses processes that were engineered 30, 40, 50 years ago? How do you adopt new technologies to those business processes?
I believe the larger change needs to happen on the management side, and to think about what makes common sense with these lines of business. We need to make those investments, but ensure we're leveraging these newer technologies.
What is your view of how the government/contractor relationship will change? Will there still be true partnerships?
I believe that the partnerships will actually move to higher-value work. What I mean by that is that if you look at a lot of spending right now, we're not addressing some of the tough issues — issues around re-engineering how these agencies work, rather than just going out and spending money on servers, routers and switches, and configuring them and upgrading them two years later. All you've done is gone through multiple cycles of Moore's Law. People profit off of that, obviously. But we want to make to sure we are engaging in work that will fundamentally change the way the government works through technology.
It's not a zero-sum game. If we don't spend money on infrastructure, will we find a game-changing way of doing it at a tenth of the cost? How do we make sure we're engaging the private-sector community to help in areas that really need work? How do we build the 21st-century Social Security Administration, or the 21st-century Health and Human Services Department, or a 21st-century Department of Defense?
What is the new administration's view of the federal enterprise architecture?
As we think of architecture in general, it’s great to have architecture that is actually being used. It's meaningless to have architecture filed away in cabinets. You could have the best document that is just sitting somewhere, yet everyone else is moving forward and implementing a completely different model.
Part of that focus needs to be on how to move forward in terms of execution and implementation, not only in terms of procuring systems, but how are we making sure that as they are being configured they are they integrated with this larger architectural vision? And as we evolve the architecture, we should evolve it to what the business side of the house [does], so it doesn't become an abstract discussion or document.
Are there any agencies doing that now that you have seen in your time in office?
We're working very closely with the Centers for Medicare and Medicaid Services, looking at the enterprise architecture there and figuring out what would be good in terms of health care and especially in the content of the kinds of investments we've made in health care IT.
You've talked about making government acquisition processes faster, but procurement is a slow process put in place deliberately to stop agencies from making rash decisions. How will you make sure such oversight remains as the procurement process speeds up?
Rules are necessary for a couple of reasons. We need rules in place that make sure we're protecting taxpayer dollars, that competition is fair, that we're being transparent, that we're giving everyone the opportunity to compete.
But I think we need to simplify. The [GSA] storefront is one model. I don't necessarily think that we need wholesale transformation right away, though we should evolve toward that. We need to rethink how to restructure the way the General Services Administration provides some of these services. How do we restructure these processes to provide easy, fast access to procurements that are out there? Everyone in government shouldn't need to have a Ph.D. in procurement. Why is it so tough? We need to do a better job of making available some of these solutions.
The schedules out there look great. But how do we change the default discussions so people are looking at the schedule first, rather than engaging in a two- to three-year process?
You gave a recent talk at the National Defense University on setting up the GSA storefront for cloud computing services. But for many government IT workers, cloud computing is still a mystery. Do you plan to offer training for these new technologies?
If you look at the tools that are available now, you do need to have a lot of background and a lot of training in everything that is involved. The intent here is to make it extremely easy and bake into that education and training. That is the consumer model. The tools we have now are designed for the government user. How do we learn from other industries?