Flagship Riverbed Observability Platform Receives Generative, Predictive And Agentic AI Infusion

Building on its major platform launch last year, Riverbed is packing its observability platform with generative, predictive and agentic AI features. The company also unveiled observability modules for UC applications, network packets and Intel Thunderbolt connected devices, Riverbed’s CEO tells CRN.

Riverbed’s flagship AI-powered observability platform has been expanded to include generative, predictive and agentic AI features, as well as a new module for measuring unified communications (UC) performance and an expanded packet capture feature that includes visibility for packet connections outside an organization’s network, the company revealed on Tuesday.

The Riverbed platform was announced last May in a launch the company called its biggest release in years. The aim of the open, AI-powered observability platform was to fill in the blind spots that exist in complex IT environments that include public cloud and remote work environments, as well as zero trust and SD-WAN architectures. At the same time, the San Francisco-based company launched the second generation of Riverbed IQ 2.0, an AIOps service that gives IT teams a chance to solve issues faster and without human intervention.

Riverbed’s observability bookings in Q1 2025 grew 102 percent year over year, marked by customer and channel partner interest in the platform, Dave Donatelli, Riverbed’s CEO, told CRN.

The one-year-old platform has helped take Riverbed from a point product company to a platform company tapping into its strength in observability. There’s room for partners to build advisory services around observability, too, said Mike Burnstein, chief growth officer for Ashburn, Va.-based solution provider and Riverbed partner RavenTek.

“They’re going to provide [customers] with access to all the stuff that they’ve had over the years as far as optimization and application enhancement, but [enterprises] also get a real-world, AI powered engine on top of it, and we see that as a big plus,” Burnstein said.

RavenTek, which primarily works with federal clients, is seeing many of its clients trying to do more with less. The updates to the platform will help reduce time to resolution for network administrators, Burnstein said.

“It will help with anticipating disruptions; it will help with self-healing environments, and it will really help with eliminating some of the antiquated ways that help desk and IT teams try to understand their environment. Even with AI, that’s just the beginning,” he said.

The platform’s expanded AI capabilities will help IT administrators managing AIOps move from reactive to predictive through three new features. The first is Riverbed IQ Assist, a generative AI assistant for AIOps that offers up intelligent, context-rich insights with minimal prompting and without long chatbot interactions. The new feature can integrate with service management platforms like ServiceNow.

The second, Riverbed Predictive AI, has been built into Riverbed IQ Ops and analyzes historical and real-time telemetry to identify early warning signals before users notice any service disruptions.

Lastly, Agentic AI, a new feature that can be enabled by the Riverbed Data Store, is built on a growing collection of task-specific agents that can be dragged and dropped into automations that IT teams design, according to Riverbed.

“[Agentic AI] is where you get agents talking to agents and this is how things really get done,” Donatelli said. “But one of the concerns customers have is that it’s their environment. They don’t necessarily want agents going around doing things that they don’t know about.”

To remedy this, Riverbed gives enterprises full control over how the AI agents work and permissions, Donatelli said.

“We will continue to build scenarios where customers can decide: ‘OK, if this happens, I want agents to fix it automatically and without my intervention,’ or to whatever degree they want it to work,” he said.

The predictive and generative AI features are now available, with agentic shipping later this quarter, the company said.

The company also unveiled the new Riverbed UC module for real-time visibility and analytics into UC applications such as Microsoft Teams, Zoom, and WebEx. The feature tracks issues such as jitter, latency, and packet loss and integrates with Riverbed Network Performance Management (NPM)+ for network and cloud monitoring and the Aternity offering for end user experience monitoring, the company said.

Donatelli said that because users frequently have issues with UC tools, the latest offering will be very popular. “There’s a lot of customer demand for this,” he added.

Building on NPM+, which was introduced last year, Packet Capture is a new SaaS-based module that provides detailed network diagnostic, including visibility for packet connections outside an organization’s network, using endpoint packet capture support for MS, Linux, and macOS systems.

“This really takes what used to be done on a very kind of bespoke hardware-based appliance, now to a SaaS product that sits at desktops and endpoints … It’s perfect for people who have clouds and zero-trust environments,” he said.

Riverbed also unveiled Aternity for Intel Thunderbolt, the company’s own connectivity standard, and Wi-Fi. The offering will give enterprises visibility across their Intel connected environments, which includes peripherals, docking stations and monitors. The offering is now generally available, Riverbed said.