Showing results for 
Search instead for 
Did you mean: 
Create Post

How Will AI Change Federal Networking?

Level 13

Omar Rafik, SolarWinds, Senior Manager, Federal Sales Engineering

Here’s an interesting article by my colleague Mav Turner about machine learning and artificial intelligence. These technologies have recently become vogue and Mav does a great job of exploring them and their impacts on federal networking.

Machine learning (ML) and artificial intelligence (AI) could very well be the next major technological advancements to change the way federal IT pros work. These technologies can provide substantial benefits to any IT shop, particularly when it comes to security, network, and application performance.

Yet, while the federal government has been funding research into AI for years, most initiatives are still in the early stages. The reason for this is preparation. It’s critically important to prepare for AI and machine learning technologies—adopting them in a planned, purposeful manner—rather than simply applying them haphazardly to current challenges and hoping for the best.

Definition and Benefits of AI

From a high-level perspective, machine learning-based technologies create and enhance algorithms that identify patterns in large sets of data. Artificial intelligence is the ability for machines to continuously learn and apply cross-domain information to make decisions and act.

For example, AI technology allows computers to automatically recognize a threat to an agency’s infrastructure, automatically respond, and automatically thwart the attack without the assistance of the IT team.

But, remember, planning and preparation are absolutely necessary before agencies dive into AI implementation.

Preparing for AI

From a technology perspective, it’s important to be sure agency data is ready for the shift. Remember, machine learning technologies learn from existing data. Most agencies have data centers full of information—some clean, some not. To prepare, the absolute first step is to clean up the agency’s data store to ensure bad data doesn’t lead to bad automatic decisions.

Next, prepare the network by implementing network automation. Network automation will allow federal IT pros to provision a large number of network elements, for example, or automatically enhance government network performance. A good network automation package will provide insights—and automated response options—for fault, availability, performance, bandwidth, configuration, and IP address management.

Finally, strongly consider integrating any information that isn’t already integrated. For example, integrating application performance data with network automation software can automatically enhance performance. In fact, this integration can go one step further. Integrating historical data will allow the system to predict an impending spike in demand and automatically increase bandwidth levels or enable the necessary computing elements to accommodate the spike. While network automation is powerful and can start your organization down the path to prepare for AI-enabled solutions, there’s a big difference between network automation and AI.


This type of preparation provides two of the most essential elements of a successful AI implementation: efficiency and visibility. If an agency’s network isn’t already as efficient as it can be—and if the different elements of the infrastructure are not already linked—advanced technologies won’t be anywhere near as effective as they could be if the pieces were already in place, ready to take the agency to the next level.

Find the full article on our partner DLT’s blog Technically Speaking.

The SolarWinds trademarks, service marks, and logos are the exclusive property of SolarWinds Worldwide, LLC or its affiliates. All other trademarks are the property of their respective owners.


It might be that AI will introduce more complexity to federal computing.  Certainly it will narrow the field of people who are able to build and understand jobs and successfully modify/correct them.

Perhaps it will introduce more vulnerabilities.  In this day of increasingly complex and frequent and subtle attempts to access systems and data without permission, perhaps the script kiddies will be able to level up and leverage increasingly complex systems for fewer dollars and discover their efforts may result in more and deeper access than ever before.  One considers how an AI might be programmed to not only penetrate a secure system without permission, but to also efficiently cover up the access and the data transfer or data manipulation.

I hope the opposite is true.  That systems become friendlier and simultaneously more secure and intuitive, opening themselves up for increasing numbers of people who can easily understand and program and use them.

Level 14

Thanks for the article.  I think the more I see the term "AI" thrown around, I wonder where the boundaries of the definition will end up.  It seems to be moving daily....


"These technologies have recently become vogue "

And Federal Networking has jumped on the Hype Train too...

I see the biggest challenge is "clean data".  With so many data sources over the decades there is no way to normalize it.  Thus it will be an ever increasing mountain to scale.  The next question is how do you define clean data vs. dirty data?  How do you clean Petabytes of dirty data?  How do you then validate is is clean and won't be dirtied again?  How do you retain original data for forensic purposes?  I think there is a bit more to this than the article suggests...again the Hype trains rolls on.

Level 13

Thanks for the Article

Level 12

Of course this data mining provides wonderful opportunities for abuse by people who think Orwell's Ninteen Eight-Four is a how-to manual instead of a dystopian warning. Then again, with so many businesses doing the same thing, government is just getting in on the act.

Level 12

What we need is the AI to clean the data for us. It isn't very intelligent otherwise, IMHO. I.E. If a human can clean it then why can't the AI?


Thanks for the article!

Level 13

We've had an AI initiative / exploration going because we have some folks in leadership that bought into the hype.  The problem is it's about 10 years too early from what we've seen.  It sounds cool, and definitely wins points with the crowd that always wants to be on the leading edge, but the reality is well short of the hype.  Of course the engineers were saying that from the get go, but they usually don't get invited along with the sales & marketing crowd so the unbridled (and unfounded) optimism prevails.

Level 11

nice article, I say turn it on and see what it does.  makes for a great movie, so it must be a good idea

Level 20

It's seem to me it's going to really really big... we're investing big time in ML/AI and also Quantum technology.

Level 14

Yes.  Management see AI as protrayed in the movies.  We techies know better but it is hard to convince those whos IQ barely reaches double figures.

Level 11

Thanks for the article.

Level 12

thanks, this is very interesting. This is an unknown argument  for me.