Be the Squirrel
Squirrels are darn resourceful. Anyone with a bird feeder can tell you that.
I mean, even Angus MacGyver has nothing on a determined squirrel. Just when you thought you cut off every possible way they can get onto your tray feeder from the ground, you catch them flying through the air from a tree over four feet away, and landing exactly where they want to. Impressive.
Just like the squirrel, federal agencies face a lot of potential obstacles to implementing a data operation, including industry competition for highly skilled staff and the inflexibility of acquiring new software tools due to the Risk Management Framework.
Photo by Vincent van Zalinge on Unsplash
The hiring situation is just a fact of life , and the RMF is very important to keeping our government networks safe. So neither is going anywhere anytime soon. I call these potential obstacles because while they may slow down the process a bit, or make you rethink how you do things, they don't necessarily need to be obstacles. In fact, they can prompt agencies to consider better ways of accomplishing their mission, and maybe even set an example for other agencies to follow.
I've seen agencies across the entire Squirrel Spectrum of Resourcefulness (SSR, for those who like acronyms). One thing that most commonly determines where they place is the mindset of the data team, set by leadership. During discussions I've had with them, most teams start out citing the issues of staffing and software already mentioned. But then, some teams start to talk about all the ways they've overcome these while others continue to talk about the problems. The more resourceful teams are usually led by resourceful and optimistic change agents who can communicate a clear vision, find creative solutions, navigate bureaucracy, push boundaries and build coalitions. They keep their eye on the ultimate goal of supporting the mission, and relentlessly employ the resources they have to achieve that. Their animal spirit is the squirrel.
As data scientists (for example), we'd all love to have a coding environment that allows us to pull in libraries when and where we need them. We want to freely use community resources like GitHub, or AWS. We'd love to have the latest and greatest development environments, and use whatever language we feel is best. We know what we need, and it's frustrating when we can't get it, and we definitely will voice our opinion on the matter. Leadership, I'm talking to you. You can either complain with us and encourage us to push back on the IT teams, or you can lead us in finding alternate ways to achieve the mission while working with IT to sort out the environment.
And there are alternate ways. I'll say the following, at the risk of receiving massive volumes of hate mail from the data science community: it is possible (although far from ideal) to build lightweight machine learning models in that program we all love to hate: Excel. Data scientists, I'm talking to you. It's true, and I'm definitely not the only person to consider doing it, see here: https://www.datasciencecentral.com/profiles/blogs/advanced-machine-learning-with-basic-excel. In fact, Google 'machine learning in excel' and you'll see articles, videos and even books on the subject. Not ideal, but possible. If you support those with national security missions, and you don't have the tools to do it, you owe it to those people to at least give this a try while you're working with IT to build your dream environment.
In summary, be the squirrel. And reach out to Cybele Data Advisory if you want more ideas on how to overcome potential obstacles to data operations.