SolarWinds is looking to leverage the power of artificial intelligence (AI) to help transform IT service management (ITSM) by integrating a purpose-built generative AI (GenAI) engine, developed using the company’s AI by Design framework, into its Service Desk offering.

Using technology including large language models (LLMs) and custom algorithms, SolarWinds AI in Service Desk helps summarize ticket histories, suggests responses for agents to use, and gives real-time advice on how to fix problems.

The company plans to expand the use of SolarWinds AI to other parts of its software, which could benefit teams working in development, security, databases and more.

By speeding up the process of fixing problems, the feature aims to improve the experience for employees throughout the organization, reducing the time they’re affected by technical issues.

“We’ve heard from our customers across nearly all of IT, including in ITSM, that they are facing increased constraints on budget, headcounts, and resource allocation,” said RJ Gazarek, principal product marketing manager for ITSM at SolarWinds.

He explained the reality is that many of these companies have not stopped growing, both in terms of annual revenue and number of employees.

However, investments in IT teams have not matched this growth, and teams are being asked to manage these much bigger companies with a similar number of resources than previous years.

“This means these IT teams now need a helping hand, including through automations, more self-service capabilities for employees, and AI capabilities that work alongside each agent as they work,” Gazarek said.

He added by having a “sidekick” that helps them get through their simpler tickets faster, IT teams can dedicate more of their time on the more complex incidents and requests that require human input and significant mental focus.

It helps teams maximize efficiency by reducing the amount of redundant and mundane tasks they are doing,” he said.

Venky Raman, senior director of product management for SolarWinds, explained the company’s AI by Design framework was formed to guide design of secure AI-driven solutions and is guided by four key principles.

Privacy and security focus on safeguarding data through measures like role-based access control, multi-factor authentication, privileged access management and anonymization techniques, ensuring that sensitive customer data does not pass through external large language models.

“Our AI engines run on top of the data platform, which ingests data from our customer environments as logs, metrics, traces, assets, alerts and change events,” Raman said. “It entails establishing access policies, implementing data classification, and ensuring data protection.”

Strong encryption measures must be implemented to safeguard data at rest and in transit, covering storage, communication, backups and replicas.

Controlling access through RBAC, MFA and PAM prevents unauthorized entry, while anonymization and pseudonymization techniques help mitigate privacy risks, removing or replacing identifiable information.

The accountability and fairness principle ensures that a human in the loop can regulate the decisions made by AI.

“Feedback and validation mechanisms will be built in, ensuring that any negative user experiences are proactively captured and addressed,” Raman said. “Regularly evaluating the models for fairness is essential for mitigating biases in the application of AI.”

He noted fairness principles strive to mitigate bias and discrimination in AI systems.

Transparency and trust are achieved through an explainability pipeline that clarifies the rationale behind AI actions, fostering continuous improvement based on customer needs.

“An AI-based experience in our applications is built on simplicity, contextual relevance and actionable insights,” Raman said.

Finally, simplicity and accessibility principles help ensure the technology integrates into user-friendly interfaces, maintaining regular workflows and user practices.

“Building trust with users on AI-driven decisions, for example, by highlighting the value or business benefit rather than just technology, is a crucial factor influencing the success of these experiences,” he said.

An example is the Anomaly-Based Alerts feature, aimed at reducing alert fatigue resulting from the noise of too many alerts.

“We strengthen existing user behavior and practices by allowing our users to build upon defined alerts with contextual anomaly detection data,” Raman said.

Techstrong TV

Click full-screen to enable volume control
Watch latest episodes and shows