As Cisco infuses its enterprise infrastructure products with AI, the technology is also infiltrating its internal platforms, directed by Cisco CIO Fletcher Previn. His focus is on developing and using AI tools that can change how Cisco and its employees do business.
Cisco is immersed in AI, looking for ways to make life and work better for employees and customers, Previn told Network World. Cisco employees are using AI tools to accelerate software development and operations, for example. Tasks that might have taken 15 hours can now be done in 15 minutes.
“We’ve spent 50 years learning to talk to machines. Now machines are learning to talk to us. The next great programming language is English—and that will reshape what roles, skills, and structures are needed inside IT,” Previn said.
[Read more Cisco news and insights]
AI is simplifying what used to be labor-intensive processes, including AI-assisted patch remediation and real-time wireless optimization. “Site surveys – we used to send people around the building to adjust the power on wireless gear – that we used to do once a year are now handled every 30 minutes with AI. These aren’t just marginal improvements—they’re orders-of-magnitude productivity gains,” Previn said.
Since becoming Cisco CIO in 2022, Previn has been charged with simplifying the technology he and his team work with as well as exponentially enhancing employee experiences. He sees as AI as well suited to both objectives.
AI can democratize access to information to deliver a “white-glove experience” once reserved for senior executives, Previn said. That might include, for example, real-time information retrieval and intelligent process execution for every employee.
“Usually, in a large company, you’ve got senior executives, and you’ve got early career hires, and it’s not financially feasible for everyone in the company to have an executive assistant, a chief of staff, an HR partner, a lawyer, a graphic designer, a communications person, a blog writer, and so on and so on. But with AI, we can actually do that at scale for everything in our company and create this kind of white-glove experience for everybody,” Previn said.
Previn singled out a Cisco agent called Circuit, an employee-facing AI assistant designed to act like a trusted coworker who has been at Cisco for decades.
“Circuit connects to a range of data sources and AI models to carry out tasks on behalf of employees, whether it’s checking calendars, initiating vacation requests or interacting with HR systems,” he said. “Instead of forcing employees to learn tools or navigate APIs, AI agents orchestrate processes across systems behind the scenes. The goal is to evolve enterprise work from searching for data to simply stating intent—like having a conversation with a knowledgeable, action-oriented colleague.”
Circuit is just one of many agents Previn said the company uses internally. But does a proliferation of agents contradict Previn’s simplification tenant?
While growth in agents may increase complexity for a time, Previn believes that will be a temporary issue. “Ultimately, as agents become more capable, the need to know where or how to complete tasks disappears. Tasks are delegated to the right AI agents, who work across systems transparently without involving the employee,” Previn said.
A user can define intent, or what they want to know or do, and they won’t have to execute the actual workflow because AI will handle the orchestration – that’s simplification, Previn said.
On the topic of AI security, “we are already almost at the point where the only thing that can defend against AI is more AI,” Previn said. Given the economics of it, “it’s become so much cheaper to be a bad actor than a good actor,” he said. “If adversaries are using these best practices, [then] we have to be sure we’re using these best practices to defend ourselves. There are a lot of apex predators out there, but Cisco is in the business of fighting apex predators.”
Cisco’s advantage is that it builds all the key components of an enterprise stack—from switches to security and observability platforms, Previn said. “I get to take advantage of products such as AppDynamics, ThousandEyes, Duo, Umbrella, which are tied into a centralized telemetry and observability framework—ingesting over 40 terabytes of data daily —and I get to take advantage of all of that,” Previn said.
“When I put all those things together, I have everything I need to build an end-to-end SOC. We also build AI-powered network infrastructure and validated blueprints that can be sold to customers, which is sort of unique to being the CIO at Cisco. And then we leverage AI on top of it to get some real AIOps, and we can react in real time to what is happening across our enterprise,” Previn said.
“So, the challenge is that everybody wants access to everything instantly, yesterday. But we’re also the group that needs to make sure that we’re doing what is compatible with our values, that does not introduce unnecessary risk to the company, and that we and our employees have a handle on what these technologies are actually doing,” Previn said.
Source:: Network World