The public sector has undergone significant change over the last year, with COVID-19 posing a unique time and challenge for the industry. But the pandemic in many ways underscored and accelerated a critical need that was already growing for the sector: the need to modernize the way that federal agencies operate and innovate.
At the heart of this idea is having the right technology infrastructure in place to fuel advanced digital innovation and to enable applications of the future, particularly around the rapidly growing potential of AI. However, the infrastructure that these organizations have relied on for decades has provided limited access to their most important resource: data.
Moving forward, pandemic or not, there is a clear opportunity for federal teams to prioritize real-time, secure access to data in order to unleash new insights, discoveries and innovations from these growing datasets.
Legacy systems that have been part of government infrastructure for decades have resulted in costly errors, security vulnerabilities and an inability to effectively serve people and organizations. While many factors can play a role here, the biggest is often the challenge of infrastructure complexity, which has plagued the public sector immensely.
This problem often comes from decentralized, disparate infrastructure that makes it difficult to have a streamlined and unified view of your data, particularly a way to access it in a fast, reliable and efficient manner. Whether it’s having massively growing datasets spread across different tiers of data storage, struggling to match the right hardware with the right software, or simply relying on a long list of different technology vendors to keep your operations running, there are often too many cooks in the kitchen to tackle data infrastructure in the right way. This is especially true in a hybrid, multi-cloud environment where it can be extremely difficult to gain visibility into all data and all applications spread across an organization.
The unique nature of the federal sector also poses a different set of challenges than other organizations may typically experience in more of an enterprise environment. For example, while the issue of scale can oftentimes be a universal obstacle for any organization in any industry, the high-stakes nature of federal programs, particularly at the defense or public health level, are paramount and makes the ability to use data effectively and efficiently to respond to situations in real-time critical.
The federal ecosystem can change easily, throwing curve balls that can make this already difficult challenge even more daunting. This is especially the case given the natural fluidity that comes with federal budgets, policies and overall priorities, which can change faster than an election cycle. Having an affordable approach that also provides flexibility is a must.
The first step toward a solution that can often be overlooked is truly recognizing the power and value of having fast access to all available data. In today’s AI-driven world, this is especially the case, with machines hungry for all obtainable resources in order to achieve the best results. And this is particularly important with applications such as simulation testing, whether it’s mapping the spread of disease or weather, or simulating aerodynamics.
But regardless of the use case, the more data you’re able to throw at these algorithms and applications, the more accurate and stronger they become. And this takes more than just “a lot” of data — it requires all of it. Organizations often assume older data does not offer any benefit in today’s world, but data that is five years old can have the same value as when it’s five minutes old.
Once this step is realized, the challenge is not over. Creating the underpinnings to actually feed these solutions with all of your data — and quickly — is a whole other matter. This is especially the case with cost considerations where limited budgets have made it nearly impossible to affordably manage and store data at scale.
However, emerging technologies such as NVMe over fabrics, storage class memory, and low-cost QLC flash are reversing a long-standing problem that has prevented organizations from leveraging a key component to having this fast access — flash storage. Traditionally a highly expensive technology, flash is becoming more affordable, giving organizations the ability to democratize their data and have near real-time access to the information they need. This frees teams from having to store large amounts of data on hard drives, and also offers increased reliability that comes from no longer depending on mechanical, moving media, which can more easily break or deteriorate. But this type of technology must also come from a supply chain that is secure, and that offers solutions that don’t put valuable data and applications at risk.
Perhaps the most overlooked consideration when it comes to technology infrastructure is people. It may seem odd to think about people when it comes to a technical topic like infrastructure, but it is important to remember both what — and who — these applications serve. There is only so much data that a researcher or analyst can work with and consume within a 40 or 60-hour workweek. And a team of people can only be so large, especially with organizations that have a more limited budget.
Having open-ended conversations with infrastructure teams is imperative to truly understand what the problem is and how a potential solution can actually be a benefit to someone. Technology infrastructure must serve and support the people that work on these applications, as well as the people that these federal agencies are trying to serve. Having technology that supports this end goal is the ultimate ingredient for success and one that cannot be forgotten or overlooked.
Future of federal
Whether it’s a stronger ability to track and predict the spread of diseases, provide more prescriptive healthcare to individuals, have new ways to respond to threats from adversaries, or tackle large issues like climate change with new and smarter approaches to energy use, the sky’s the limit for federal organizations.
Regardless of the different challenges and opportunities that federal agencies and their partners may face, the ones that remove the bottlenecks to using data effectively and efficiently will be the ones that better meet the needs of the people and communities they serve.
Randy Hayes is vice president of Public Sector at VAST Data.