Virtana has added Agentic Observability Enterprise Management to its observability platform, allowing its AI agents to pull agents from other vendors into work on IT incidents. Any outside agent that supports Model Context Protocol or the open Agent2Agent protocol can join an investigation, including agents already bundled with software a customer owns. The move targets outages that spread across systems from several vendors, where no single built-in agent sees the full picture, which matters for cutting the manual work of joining up findings.
How cross-vendor agent response works
The release addresses how enterprises have adopted agents so far. Most agents arrive built into one product, such as an IT service management system, and stay inside that boundary. When a failure crosses several vendors, engineers still collect and reconcile what each agent reports. Virtana positions its own agent as the coordinator that spots trouble and invites other agents into a shared investigation instead of leaving that stitching to staff.
The mechanism relies on two open protocols and two data feeds Virtana built earlier. Outside agents have been able to query Virtana monitoring data since February, when the company released its System-Aware MCP Server with more than 25,000 real-time metrics. Called-in agents also receive a shared view of application-to-infrastructure links from the System Dependency Graph, introduced at the same time. A month later, Virtana extended monitoring to AI workloads, widening the telemetry available to the group.
Virtana illustrates the flow with a slow application triggering a request for help. A third-party agent notices the slowdown and locates the Virtana root cause analysis agent through Agent2Agent. That agent traces the issue to a database bottleneck that appeared after a recent software deployment. Follow-up questions can return to the root cause agent before other specialized agents are lined up for the fix, after which Virtana software confirms the application is back to normal.
What this means for incident operations
For companies running multi-vendor estates, the practical change is faster assembly of the right expertise around an outage. A conventional IT war room takes hours to staff and run, while a team of agents can get through the same work in milliseconds, according to Chief Product Officer Amitkumar Rathi. Small teams gain coverage they could not staff around the clock, while large enterprises can run parallel investigations without pulling engineers off other systems for the first triage.
Autonomy stays bounded by customer policy, and that boundary shapes adoption risk. The agent teamwork applies only to incident types a customer has defined in advance. Customers seeking more control can keep fixes with engineers, require sign-off on each agent action, limit which outside agents may join, and cap how many tokens those agents may consume. Every agent action is written to an audit log, giving compliance teams a record to review.
The marker to watch is whether customers expand the list of pre-approved incident types and let agents execute fixes without step-by-step sign-off. Virtana, known as Virtual Instruments until a 2019 rebrand, raised $73 million in 2022 from investors including Atalaya Capital Management LP and Benhamou Global Ventures. Broader sign-off-free use across production systems would show that cross-vendor agent teams have moved from pilot to standard operations.
