Becky Smith is Principal at SoundWay Consulting Inc., and has more than 25 years of experience supporting the Department of Defense (DoD) and Intelligence Community (IC) as a government contractor.
Smith spoke with WashingtonExec about her role at SoundWay, open source, Big Data analytics, the “data gap,” women in the industry, and more.
WashingtonExec: Could you start off by telling us a little about your background and your current role at SoundWay?
Becky Smith: I’ve been in IT since the early ’80s, and the Intelligence Community and DOD intelligence for about fifteen years or more now. My emphasis in graduate school was databases and searching disparate databases and from there, I transitioned into data standards and Intelligence Community standards in particular. My role at SoundWay is Technology Lead and I am in charge of the practice that is doing the big data and cloud work right now. SoundWay is a small woman-owned, service disabled veteran business.
WashingtonExec: What is the largest issue you have faced in trying to start these big data initiatives in the Intelligence Community and Defense Community?
Becky Smith: The Intelligence Community is trying to move to IC ITE, (Intelligence Community Information Technology Environment). The point of IC ITE is to achieve IT efficiencies and reduce redundant systems across the Intelligence Agencies, and also take advantage of not only saving money but leveraging new technologies and common standards. For example, if you co-locate your data and tag it in a common way, you can do things like run analytics across all of your data. Right now, the agencies have separate implementations for all of this.
“The security policies aren’t the same across the IC. All of those things make it hard to share and understand data. There is also a cultural predisposition to want to hold on to your data.”
WashingtonExec: Do you think the Intelligence Community (IC) will embrace open source now or do we still have a long way to go?
Becky Smith: I’m seeing a lot of embracing of Open Source software right now. We’ve almost moved away from COTS Solutions and we are using a lot of Open Source software, particularly in the cloud implementations that are going to be a part of IC ITE. In my mind that can be a good thing but we tend to think that Open Source is going to be cheaper and that’s not always the case. We’re favoring Open Source and I think we need to somehow settle somewhere in the middle and reach a balance between what makes sense to buy products for and what we can build with open source. A lot of COTS products have years of intellectual capitol devoted to a particular solution and that would be hard to reproduce with Open Source software.
WashingtonExec: Where are we in the process of gathering, coding, and then implementing big data analytics? Do we have a long way to go?
Becky Smith: I think being able to run analytics over big data gives us opportunities to mine information in open data sources like we never have before. The balance is going to be how to triage data or keep it around enough so you have it when you need it, but you’re not wasting cycles on data you don’t need to be processing. I think Big Data implementations are evolving. As far as analytics, I think we are just beginning. Ultimately it would be great if analysts could set up and run their own analytics in an intuitive way but we are a ways off from that.
WashingtonExec: Do you think security rules and regulations need to be updated?
Becky Smith: I think we are hindered by our security policy right now, or our disparate interpretations of security policies. We have memorandums of agreement on data sources. We have legal restrictions on data. We have classifications on data. All these things result in confusing/competing data access and handling rules. The rules are not consistent across the IC and that is a shame because if we can mix that data, co-locate that data, and share that data we have the opportunity to find out a lot more than we know now with the use of analytics and/or different eyes on the data. Until we can crack that nut we’re not really going to be able to attain the level information sharing and analytics that we want to achieve.
WashingtonExec: Do you believe in the “data gap”?
Becky Smith: I think we have to change the paradigm where we have to manually process all of the data. We can’t possibly have human eyes on all of the data so we need to take advantage of the analytics and machine processing to triage what we can and filter what we can. Right now those analytics pretty much require data scientists with vast experience in data and in statistics. These people right now are hard to find.
WashingtonExec: Are you involved at all with predictive analytics and do you see this as a natural progression of this kind of analysis in the IC community or just more of a fundamental shift?
Becky Smith: Right now my focus is helping the agencies get their data into the government cloud so that we can run analytics on it, but we are beginning to have the discussions across the agencies about what they are doing with analytics, what a framework for sharing those among other agencies would be and getting to the point where we can dynamically set them up and share them and have them be interoperable. That’s an evolving process.
WashingtonExec: Who is someone you admire?
Becky Smith: I’ve had a lot of government customers that do this day in and day out just because they care about their country so much. They are often not highly paid. They don’t always get the respect that they deserve. We have some people who have been dedicated for decades to helping our analysts be better, helping with issues of 9/11 and I have a lot of respect for them. I have a lot of respect for women in this industry that are senior leaders.
“This traditionally was a male dominated industry and culture. The woman that has been able to rise to the top and drive where we are going should be commended.”
WashingtonExec: As someone who has been in this industry for the last 25 years, have you seen a shift with more women in the industry?
Becky Smith: I’ve seen a lot of change. There is no woman my age in this industry who hasn’t been talked over in meetings or had to be the “booth babe.” I’m seeing less and less of that. My daughter and I both had NROTC scholarships and her experience was vastly different than mine was. I have seen great changes and I think there are going to be greater opportunities for women. I look up to women who have put up with all of this stuff and are making a difference.
WashingtonExec: I’ve hearing industry executives complain that there are now too many protests.
Becky Smith: I’m seeing that too. Everything is taking so long to be awarded – there are a lot more protests. As a small fledgling company, when we write a proposal there are not a whole lot of resources to draw on so it is taking a lot of corporate resources and a lot of cost that is more significant to us than it might be to a big integrator. We’d like to see more set-asides. It’s very discouraging to put in all the work and have the awards delayed/canceled/protested. We’d like to see a more streamlined acquisition process of course. I think everybody would but it is particularly hard for small companies to pick what’s worth going for and what’s not.
WashingtonExec: Some technologists have said that the largest issue with big data is not in regards to technological advancement, but is a cultural one, with getting leaders comfortable with “releasing their data,” do you agree or disagree?
Becky Smith: I agree that this continues to be an issue in the Intelligence Community though it is improving. There is now data the agencies freely share; it is harder to obtain access to mission specific data. Most of this issue is due is lack of trust that the data will be protected properly. That is why standardizing the security policies is so key; then everyone will understand how the data will be stored and shared. Until we can do that, we will not be able to reap all the promise of Big Data.
WashingtonExec: How do you identify what is important data and what is “clutter”?
Becky Smith: I was reading a blog posts by analysts; it was making the point that much of the data is noise and we need to be able to find the signal within the mass of data. This is a difficult problem because what is not important today, may be very important tomorrow. Our data needs change and evolve. So we need to be able to keep all the data around to refer back to but we need to be able to determine meaningful patterns in the data and filter what we currently need, as well.
WashingtonExec: Define big data in one sentence or less:
Becky Smith: To use the Standard answer, the 5 V’s: volume, velocity, variety, variability and value.