How Cisco’s CIO Uses AI And Tech To Expand Company Culture – Forbes
ByMegan Poinski,
Forbes Staff.
For the last four years, Fletcher Previn has been a senior vice president and CIO at Cisco, using technology to both expand and enhance company culture for its more than 100,000 employees and create new opportunities and efficiencies through AI. Before coming to Cisco, he spent almost four years as CIO at another big tech company: IBM. I talked to Previn about how he uses his position to move Cisco’s culture and technology forward.
This conversation has been edited for length, clarity and continuity. It was excerpted in the Forbes CIO newsletter.
What made you want to work at Cisco?
Previn: I was drawn to the mission of Cisco, connecting everyone and everything and building out the global infrastructure of the future. It’s one of those companies that plays a critical role in defining what kind of world we’re going to live in, and enabling access for people. I had worked a fair amount with Cisco as a customer over the years, and I felt like I knew Cisco as well as you can know a place without actually working there. I liked the culture and mission, and that’s really what drew me here.
What is it like being a CIO at a tech company, and, more specifically, a B2B tech company?
Working in tech is always interesting, always changing, always exciting. All of those things are doubly true when you’re at a company that is in the business of inventing that technology. That is the excitement. The challenge is that you get to do things that you’re customer zero of—which means there are zero other people that you can consult about what is the right way to do this. You get to really be involved in the product development side of the house, and generating requirements and testing things before we sell them to customers. So it is both exciting and a unique set of challenges.
I had a friend once tell me doing IT at an IT company is like being a brain surgeon if they invented a new brain every 18 months. It’s not our goal to just be competitive in the market. We’re Cisco, so we should be doing things that people have never seen before. That means you’re always learning, you’re always growing your skillset and doing things. Even if it’s a thing you’ve not done before, you’re growing.
Cisco Senior Vice President and CIO Fletcher Previn.
You talked about the culture at Cisco and how you find it important. How do you bring culture to your role as CIO, and how does it manifest its way through information?
We spend a lot of time thinking about technology as a driver of culture change and the idea that your culture is really the truly unique thing that you have. Your culture is really a function of how work gets done, and the shortest distance to engaging employees is what’s in their hand or what’s on their desks. So how well we’re delivering it to the workforce at Cisco is not trivial. It’s actually the core to creating a culture of high performance.
Philosophically, I’m looking to work with people who are kind, passionate about their craft and believe in our purpose. If we surround ourselves with those kinds of people, we’re going to be successful in whatever problem of the day we happen to be working on.
From a CIO standpoint or an IT standpoint, [positive culture is] really focusing on those experiences that touch all employees, taking seriously where there’s friction in the environment, and prioritizing getting after those things to have these moments that delight and surprise people. It’s very easy when you are leading an IT organization to spend the bulk of your time on some of the larger, complex, maybe more obvious things around ERP, or platform consolidation projects. But it is these quality of life issues for everyone who works at the company that form their opinion of the IT department, and can lead people to either believe that we are working in service of them—it’s a very servant function; we serve those who serve others—or that we’re not valuing their time and we’re not prioritizing.
The quality of IT is a daily reflection of what the organization thinks and feels about its people, and so we have to take that really seriously. If there are these things that are dissatisfiers in the environment, how often do people experience this? How bad is this experience, and how important are the criteria that I use to prioritize my backlog—whether it’s onboarding or expense reporting or travel or device setup. These things that touch everybody, we have to take those very seriously.
As you form your relationship with the business, if you want them to trust you to lead large complex business transformation, it is hard to do that when the help desk is bad and the laptop setup isn’t good and the Wi-Fi is poor. That leads people to go: You can’t even get my laptop. Why would I entrust you with this critical business function? They are interrelated.
What are some of the non-AI projects and initiatives that you’ve done at Cisco?
One of the projects I’m especially proud of is what we call the hybrid worker bundle. When I first came to Cisco and looked at how we were provisioning people for hybrid work, it was working, but I had the luxury of being at Cisco. If you take a complex problem like hybrid work where you have to solve for cybersecurity, remote access, collaboration, device management, observability, application performance, things happening on networks at the last mile that you don’t own or manage in peoples’ homes—effectively, everyone’s home had become a branch office to some degree during Covid, and if you go from having X number of offices to having a 100,000 offices, that’s a sort of different proposition.
Cisco’s got the portfolio to actually solve that end to end. We have all those parts of technology, and hardware, software and services. I said, could we go through our catalog and put all these things together into a highly designed, highly integrated consumer-like ‘branch office in a box’ and send that to people? That’s what we started to do. It’s almost like a layer cake, where you’ve got Cisco Wi-Fi and a firewall, your laptop if you’re a new hire, your employee badge if you’re a new hire, a keyboard and a track pad. Then it's all bundled in this box that has messaging about why you should be really glad that you’re now at Cisco, and how to engage with us if you need any help. Basically, the instructions are: Step one, plug it in; Step two, turn it on; Step three, there is no step three because the whole thing provisions itself over the cloud. That really, I think, was one of those moments for employees where it signaled to people that we’re here working for them—not doing things to them—and trying to guarantee that they have a great Cisco work experience, whether they’re in the office, on the road, at home, or anywhere in between.
Cisco is evolving from a history of selling hardware to partners, to selling things as a subscription and outcomes as a subscription. That, as you might imagine, requires a lot of plumbing changes to things like CRM in terms of how you go and compensate salespeople. Licensing simplification is a big initiative at Cisco right now, to make the way we license our products simpler for customers and partners. Having the ERP and supply chain and licensing systems set up to be geared less towards a pallet of hardware and more towards selling you some outcome for a few dollars a month, which is a very different sort of exercise. That’s another big focus for me at the moment.
And then a lot of these employee touch points. Evolving that intranet to become the digital equivalent of a friend that has worked at Cisco for 40 years and just knows the answer to everything.
How about some of the AI initiatives that you are working on?
We have an internal digital assistant which takes a large language model, runs in a secure instance in our Cisco data center, and then we augment that through RAG and other techniques to know things about Cisco that public versions of language models don’t know, so that we can enable salespeople, software developers, general employees, customers and partners. You can go there and ask it things: Help me pitch this particular product to this particular kind of customer. How do I configure this piece of Cisco equipment? When does this piece of Cisco equipment go end of life and which one replaces it? How do I book travel? What’s the Cisco policy for this or that? What’s the one sentence in this 300-page document that is relevant to the thing I’m looking for? We’ve been steadily evolving those capabilities.
In order to avoid the kind of chatbot sprawl that occurred in the past, where everybody made chatbots and then you got hundreds and you don’t know where to go. It all comes together in this one place, which we instantiate through Webex or our intranet or through mobile app.
A big focus is thinking about how can AI be a force multiplier for productivity. We’ve got close to 100,000 employees here. Any number times 100,000 is a big number. If you think about the average human being liv[ing] to be about 77 years old, and assume a 40-hour work week, that translates roughly to 144 months of your life you’re going to spend working. About 60 of those months are going to be spent in meetings, and about another 44 months of those are going to be doing administrivia that’s not really core to the thing you were hired to do. So for a lot of people, approximately 75% of your work life is going to be doing something other than what you said you wanted to be when you grow up. That’s not good.
That’s a huge opportunity for AI to come in and allow us to focus on the things that are rewarding, that benefit from a human being doing them, and are the things that you feel fulfilled doing and were actually hired to do—and offload everything else to some sort of AI digital twin of yourself. That’s really the path that we’re on with these things: to move from finding and summarizing, to acting on your behalf, to allowing you to be significantly more productive.
And of course, there’s a lot going on with infrastructure operations. We are an infrastructure company. I run a lot of infrastructure. I run a lot of data centers, and we’re taking advantage of a lot of the new AI features that Cisco’s putting into its products. We used to go [to] all the Cisco offices around the world and do a Wi-Fi survey about once a year and just make sure there’s no dead spots or needs tweaking. That’s something that AI does for us now automatically every 30 minutes. You’ve got that at scale across the board in all of our internal data center operations.
What is your philosophy about using AI?
I think of AI as a teammate, not a team replacement, and as a really significant enabler for productivity. People have spent a lot of time and effort getting good at figuring out how to program computers. And now, what has changed is computers have learned how to talk to us. That has some profound impacts on things—not the least of which is the next great programming language is probably English, which means I should expect to be able to get a lot more productivity per software developer than before. There will be growth of new types of IT and technology professions around things like prompt engineer, flow engineer, people who really understand end-to-end business processes. The art of AI is going to be learning and developing the skill of what are the right questions to ask.
That will change the type of training that we do, the profile of people that we’re looking for. So we spend a lot of time thinking about what is the IT organization of several years from now.
Some near-term OKRs we’ve set for ourselves is that 50% of everything we do in Cisco should be assisted by AI. We should be trying to achieve a 50% improvement in the employee experience by reducing the measurable friction in the top 150 tasks that employees have to complete. We measure that in a number of ways with NPS, task completion and user interviews, and developed something called a task friction index. If you’re an application owner, you can go in and look and see the balanced scorecard letter grade of the application you own. It doesn’t matter how good or bad it is. The point is, for the period of time that you’re responsible for it, it needs to improve.
How are you doing on those AI use goals?
That is going to be a longer-than-2025 goal because the pie also keeps getting bigger. We’re doing more, but we’re making progress on it. One of the things I’ve done is create as a direct report to myself an organization responsible for design and user experience, and I act as the product owner on that team’s backlog and prioritize their backlog. I make certain equity decisions about how that team’s time is most effectively spent, and really try to prioritize the things that are touching all employees and make those things better for people. When I say design and experience team, I mean content research, analytics, UI, UX, visual design—the full stack of things you need to understand what problem you’re really solving. Everything from the choice of words on the page to the flow and layout and click-through to the visual design.
We’re making progress. We have a lot of data that shows that we are getting more productive, that our infrastructure and operations and cybersecurity activities are really being helped by the AI tooling in a pretty significant way, where we can do a lot more. Every IT department has the same forever problem, which is you’ve been asked to do 10x more than you can. The supply and demand is always a challenge. AI is really helping us with that problem and [we can] be more efficient at writing custom software, be more efficient in our IT operations, be more productive for general employees for ourselves in the IT department, for engineers and software developers, for salespeople and for customers and partners, to make it easier for them to do business with us.
What else do you have planned in the near-term that you’re working on?
There’s a significant relationship between your data strategy and your AI strategy and really treating data like a product, where there’s a product development roadmap with feature releases and functionality releases. [We want to have] a data strategy that allows AI to exploit that data, and for other parts of the company to be able to tap into that AI and data framework so that everyone is not having to develop things from zero, but can plug in whatever language model they happen to want to use and be able to take advantage of the security that you’ve built, the identity and access management, the internal knowledge of things that the public internet doesn’t know. That’s really the strategy: to be the centralized AI infrastructure and framework provider for the rest of the company.
One of the things that we’ve been doing is building a fair amount of our own internal AI infrastructure. We’re using Nvidia GPUs, but Cisco Ethernet Fabric instead of InfiniBand or the Nvidia DGX SuperPODs. Our own compute, our own storage, our own software to really be able to have a Cisco-on-Cisco AI story, then use that infrastructure to do real world workloads, whether it’s noise cancellation for Webex [or figuring out] what is the next likely best action for a salesperson with a customer.
We’ve built about a 1,000-GPU cluster so far, and we’ll build an equally larger or greater additional set of infrastructure this year. We’re the centralized AI infrastructure provider across Cisco.
You’ve been in the CIO role at both Cisco and IBM. How has it evolved, and how do you see it changing in the future?
I think the role of the CIO is getting steadily more technical, and every business is basically a digital business. Every company is a tech company, and all transformation paths go through IT. You really become [a] critical path for enabling the business and the strategy that they want to execute.
To do that well, people want things to be simpler. Their perception of what is a good experience is being informed by the experiences they’re having in their personal life. The architecture and the world and the landscape is actually getting quite a bit more complicated. So you have those two opposing forces: Complicated is easy, simple is hard. How do I make this complex thing easy and simple to consume and understand and use? I think that that role is becoming more technical, and the era of the generalist just passing through it is numbered. That requires pretty deep subject matter expertise to do it well.
I would also say the speed at which change is happening in our industry is a little different than previous technology trends. You would typically invent something or develop something, put it out to the world, collect feedback from people, iterate on it, put a new version of it out. There’s a certain cadence to that that takes months or years. You now have a world where AI is essentially self-optimizing in a continuous loop. There’s this hockey stick of innovation happening, where things are changing much more rapidly than they have in the past. The organizations that can figure out how to take advantage of AI effectively, while creating an environment where talented people want to work, and where you can attract and retain the best people, are going to be the organizations that win in this world.
source
This is a newsfeed from leading technology publications. No additional editorial review has been performed before posting.
