Operating an enterprise IT environment is fraught with challenges. Increased amounts of data, cumbersome manual monitoring processes and limited human staff are common hurdles. While these may seem insurmountable to agencies, addressing these challenges is possible, and it all starts with understanding the problems, skillfully applying artificial intelligence solutions — and building well-founded confidence between humans and machines.
No one knows this better than Leidos’ Director of AIOps Kevin Chin, who brings a wealth of experience working alongside government customers to modernize legacy tools and techniques. Educated as a software engineer, Chin now applies his experience as a software developer, systems engineer, and technical architect to deliver Artificial Intelligence and Machine Learning (AI/ML) to bear on IT operations.
AI and ML pose tangible solutions to the challenge of securely operating today’s agency IT environments. But where should agencies start and how can they adopt new AI/ML technologies that improve resiliency and increase human efficiency while also reducing business disruption? Here, Chin offers insight into how Leidos’ Trusted AI approach to artificial intelligence enables federal agencies to deliver AIOps automations that build on existing investments.
AI as a ‘Staff Force Multiplier’ and Silo Buster
As the world becomes increasingly globally connected, there continues to be an exponential growth in data. For example, an expanding number of endpoints like smartphones, tablets and the Internet of Things (IoT) have introduced large amounts of data. That deluge of information requires more robust mechanisms to analyze, but oftentimes that analysis is performed manually. Additionally, Chin notes that only a subset of collected data is used, and IT staff can find themselves stretched impossibly thin to analyze that data and perform correlation to identify relationships within the data.
“That’s not sustainable,” he says. “Artificial intelligence is a game-changer. It acts as a human IT staff force multiplier.”
AI, Chin says, can increase service availability and reduce business disruptions. It’s better suited to handle the data explosion to automate triage and identify remediation. It can also assist with prioritizing alert notifications, which can overwhelm a human operator. As agency operations further modernize beyond cloud computing, AI can continue to fuel automation for efficient operations. One way is to bring together data sets from different domains and locations — an increasingly necessary capability given the depth of data agencies have today and the attempts across the government landscape to operationalize it efficiently. The Department of Homeland Security, for example, can leverage Leidos’ AIOps to share data more efficiently between its law enforcement and civilian defense organizations while still maintaining a federated operating model.
“We’re breaking down the silos,” Chin says. “Now, we are able to see how different services impact one another while still supporting that federated model and giving different branches their autonomy to implement IT services.”
Differentiating Through Trusted AI
But it’s not enough to simply have AI/ML parse and triage data, Chin says humans must also be able to understand and trust the analytic result. To achieve that trust, Leidos’ Trusted AI framework provides the transparency, explainability, and observability to the analytic consumer. Chin notes that Leidos views AIOps through a foundational focus on ‘Trusted AI’. “That’s the Leidos differentiation.”
“Many vendors have AIOps capabilities, but not everyone evaluates data across domains to interpret how one data signal impacts the other in a different domain.”
Leidos has been intentional about differentiating its AIOps approach. While many vendors use data point anomaly detection to look for data point outliers to flag anomalies, Leidos pulls in data from multiple domains and applies advanced mathematics, probability distribution, and statistical analysis to examine their relationships and how one can impact the other.
In one example of applying AI internally, Leidos developed an analytic around Microsoft 365 managed services to infer when a potential service disruption will occur. It’s now used within the company and able to be offered to external customers as one more tool to use. This is especially impactful as IT organizations considering adopting more “as-a-service” managed offerings.
Speed, Security and Scale
Leidos delivers AIOps capabilities with speed, security, and scale. All AIOps capabilities are developed with a security-first mindset, resulting in the ability to deploy into environments that must meet IL6 security requirements. Combining Leidos’ experience operating its own IT enterprise with delivering solutions across varied customers enables the team to understand similar problems and issues that AI/ML can help solve. This results in ready-to-deploy analytics that can be customized to a use case.
Chin and his team are passionate about working with decision-makers to address specific needs and identifying customer-intimate use cases to mature all areas of Trusted AI. Whether its complying with Zero Trust executive order initiatives or modernizing cybersecurity posture, the Leidos team is excited to offer its AIOps and Trusted AI capabilities to support an agency’s modernization initiatives.
“We are a project, not a product,” Chin says. “We’re not here to replace vendors. We’re here to augment vendor capability by delivering use cases and known problems across the enterprises.”
Learn more about how Leidos can help your agency make use of data, secure the enterprise, and build trust between humans and machines via Trusted AI.
This content was produced by GovExec’s Studio 2G and made possible by our sponsor. The editorial staff of NextGov was not involved in its preparation.