SolarWinds launches Avior AI agents to automate network operations
Network World ·
SolarWinds this week introduced Avior, an agentic AI system, to help network teams automate operational tasks, reduce alert noise, and improve infrastructure management while letting administrators maintain control over autonomous actions. Avior builds on the company’s existing AI framework by adding goal-oriented agents that can execute tasks, not just provide recommendations. The technology works across SolarWinds’ on-premises and SaaS products, using operational data and existing management tools to resolve issues across enterprise environments. “We’ve been on this AI journey for quite some time over the past four or five years,” says Cullen Childress , interim CEO and chief product officer at SolarWinds . He says Avior extends AI capabilities with agents that can execute tasks autonomously to free up network engineers to focus on higher-level work. The first Avior offering is a network engineering agent that can automate alert management, configuration automation, compliance checks, and incident remediation. “We started there because [the network is] kind of our bread and butter,” Childress explains. “Many of our customers, a large portion of our customers, are on our network infrastructure systems monitoring side of our portfolio.” The network-focused agent can correlate alerts into actionable incidents, identify configuration drift, audit devices against compliance policies, and manage configuration backups, according to SolarWinds. The agent is also able to track capacity, monitor infrastructure utilization, and identify obsolete assets. Childress explains that SolarWinds designed Avior to give network managers flexibility over how much autonomy the agents are allowed to have. For instance, agents identify issues and recommend corrective actions in assistive mode. A human-in-the-loop mode requires an admin to approve actions before the agents can make changes. And a fully autonomous mode lets agents complete tasks independently within customer-defined controls and report the results back to the network admin. “Nobody’s going to let these agents just go do stuff unless they feel like they’re in control,” Childress adds. “Everybody has a different level of conservatism.” For instance, network managers might allow agents to automatically address repetitive alerts to reduce the noise but require approval prior to an agent making a network configuration change. Enterprises might also enable autonomous storage optimization in staging environments, while maintaining more control in production environments. SolarWinds also included several safeguards in Avior to give network teams control over autonomous agents: Risk-based guardrails: Require human approval for potentially dangerous actions. Audit trails: Track agent activity, decisions, and changes. Automated rollback: Support reversing changes when necessary. Data protection: Mask customer data and restrict access based on user permissions. Least-privilege access: Limit agents to the tools and permissions needed to complete assigned tasks. The Avior agents can also draw on telemetry, topology, configuration, and other operational data available in SolarWinds products. Avior is a separate agentic AI layer that connects with SolarWinds’ products, including SolarWinds Observability . Avior taps into their monitoring capabilities, operational data, and knowledge of the environment to inform the AI agents . Administrators manage the agents and their activities through Avior. For instance, an enterprise using SolarWinds Observability can activate Avior’s network engineering agent. “The agent uses all of the tooling and the capabilities that are inside of that product, just like the human would,” Childress says. Avior also works with third-party systems such as ServiceNow to execute workflows, and Childress says Avior is intended to reduce repetitive tasks and let network engineers concentrate on more strategic initiatives. “Our aim is to take the day-to-day tactical toil … off their plate,” Childress says. Rather than replacing IT professionals, SolarWinds envisions the agents extending the capabilities of existing teams. “It’s not just about having an assistant or a co-pilot that’s giving me advice,” Childress says. “This thing can give me advice, but it can also do things on my behalf upon my instruction.” Avior’s network engineering agent is available now through a guided onboarding program. SolarWinds expects to introduce a database administration agent later in Q4, followed by IT administration, SRE, and security analyst agents across late Q4 and early 2027.
SolarWinds this week introduced Avior, an agentic AI system, to help network teams automate operational tasks, reduce alert noise, and improve infrastructure management while letting administrators maintain control over autonomous actions. Avior builds on the company’s existing AI framework by adding goal-oriented agents that can execute tasks, not just provide recommendations. The technology works across SolarWinds’ on-premises and SaaS products, using operational data and existing management tools to resolve issues across enterprise environments. “We’ve been on this AI journey for quite some time over the past four or five years,” says Cullen Childress , interim CEO and chief product officer at SolarWinds . He says Avior extends AI capabilities with agents that can execute tasks autonomously to free up network engineers to focus on higher-level work. The first Avior offering is a network engineering agent that can automate alert management, configuration automation, compliance checks, and incident remediation. “We started there because [the network is] kind of our bread and butter,” Childress explains. “Many of our customers, a large portion of our customers, are on our network infrastructure systems monitoring side of our portfolio.” The network-focused agent can correlate alerts into actionable incidents, identify configuration drift, audit devices against compliance policies, and manage configuration backups, according to SolarWinds. The agent is also able to track capacity, monitor infrastructure utilization, and identify obsolete assets. Childress explains that SolarWinds designed Avior to give network managers flexibility over how much autonomy the agents are allowed to have. For instance, agents identify issues and recommend corrective actions in assistive mode. A human-in-the-loop mode requires an admin to approve actions before the agents can make changes. And a fully autonomous mode lets agents complete tasks independently within customer-defined controls and report the results back to the network admin. “Nobody’s going to let these agents just go do stuff unless they feel like they’re in control,” Childress adds. “Everybody has a different level of conservatism.” For instance, network managers might allow agents to automatically address repetitive alerts to reduce the noise but require approval prior to an agent making a network configuration change. Enterprises might also enable autonomous storage optimization in staging environments, while maintaining more control in production environments. SolarWinds also included several safeguards in Avior to give network teams control over autonomous agents: Risk-based guardrails: Require human approval for potentially dangerous actions. Audit trails: Track agent activity, decisions, and changes. Automated rollback: Support reversing changes when necessary. Data protection: Mask customer data and restrict access based on user permissions. Least-privilege access: Limit agents to the tools and permissions needed to complete assigned tasks. The Avior agents can also draw on telemetry, topology, configuration, and other operational data available in SolarWinds products. Avior is a separate agentic AI layer that connects with SolarWinds’ products, including SolarWinds Observability . Avior taps into their monitoring capabilities, operational data, and knowledge of the environment to inform the AI agents . Administrators manage the agents and their activities through Avior. For instance, an enterprise using SolarWinds Observability can activate Avior’s network engineering agent. “The agent uses all of the tooling and the capabilities that are inside of that product, just like the human would,” Childress says. Avior also works with third-party systems such as ServiceNow to execute workflows, and Childress says Avior is intended to reduce repetitive tasks and let network engineers concentrate on more strategic initiatives. “Our aim is to take the day-to-day tactical toil … off their plate,” Childress says. Rather than replacing IT professionals, SolarWinds envisions the agents extending the capabilities of existing teams. “It’s not just about having an assistant or a co-pilot that’s giving me advice,” Childress says. “This thing can give me advice, but it can also do things on my behalf upon my instruction.” Avior’s network engineering agent is available now through a guided onboarding program. SolarWinds expects to introduce a database administration agent later in Q4, followed by IT administration, SRE, and security analyst agents across late Q4 and early 2027.