Cisco is best known for the networks that keep the world connected, and in the AI era, that foundation matters more than ever. Increasingly, the company’s story is about what the network makes possible: security, observability, and AI that are becoming woven into the fabric itself. Nowhere is that more visible than in customer experience, where Cisco is rethinking what it means to support customers globally.
Jyotsna Khandekar sits at the center of that work. As a Director of Product Management, she leads the customer-facing digital platforms and AI initiatives that serve millions of users, including Cisco Community and the systems that surface the right insights to the right customers at the right time. Over 17 years in technology, spanning go-to-market, partner commerce, and now customer experience, she’s built a rare, end-to-end view of how product, revenue, and customer relationships all connect.
We sat down with Jyotsna to talk about where AI is genuinely improving customer experience, what it actually takes to build trust in AI systems, and why she believes the next real competitive edge isn’t your product or your data; it’s how fast your organization can turn customer understanding into action.
Execs In The Know (EITK): AI is transforming customer experiences across industries. What opportunities excite you most, and where do you see organizations getting it wrong?
Jyotsna: What excites me most is the shift AI makes possible in customer experience — moving from reactive to proactive, predictive, and personalized. For most of our support history, we waited for a customer to hit a wall and come to us. AI lets us flip that: anticipate what a customer needs and meet them with the right help at the right moment, sometimes before they’ve even had to ask. That’s a fundamentally better experience, and we’re only at the beginning of it.
Where organizations get it wrong is in starting with technology rather than the customer. There’s a lot of pressure right now to “have an AI story,” which pushes teams to adopt AI first and then hunt for a problem to attach it to. The ones who get it right do the reverse; they start with a real customer problem and ask whether AI is genuinely the best way to solve it. The other trap is optimizing for internal efficiency at the customer’s expense, measuring success by what the company saves rather than the value the customer actually gains.
EITK: Trust has become a critical topic in the age of AI. What does “trustworthy AI” mean to you, and how should organizations build trust with customers while adopting new technologies?
Jyotsna: To me, trustworthy AI comes down to three things: visibility, governance, and safety.
Visibility means you can observe which model handled a request and what data and actions were involved, audit the decisions afterward, and explain the output with clear reasoning and sources. If you can’t see what the system did and why, you can’t trust it.
Governance means real human oversight, with clear conditions for when a person can step in and when they should.
And safety means guardrails: monitoring for abnormal behavior, setting stop conditions, and hardening against attacks. When you can account for all three, and put that visibility in the customer’s hands, the AI earns trust.
Working at Cisco shapes how I see this: trust in the AI era depends on the full stack working together — the network that carries it, the security that protects it, and the observability that lets you see what’s actually happening end-to-end. That’s how you build trust with customers, too: by showing them the system is observable, governed, and safe, not just telling them.
EITK: What’s one AI myth you’d like to debunk?
Jyotsna: One myth I’d debunk is that keeping a human in the loop is a weakness, a sign your AI isn’t good enough yet. I see it the opposite way: it’s a sign of maturity, because it means you understand the stakes. Autonomy without boundaries isn’t innovation, it’s risk.
The goal isn’t to remove people as fast as possible; it’s to know exactly where human judgment matters most and design for it.
EITK: What’s one CX metric you pay closest attention to?
Jyotsna: The one I watch closest is adoption depth — how much of what a customer already owns they’re actually using. I care about it because it’s a leading indicator, not a lagging one: a renewal tells you what has already happened, while adoption depth tells you, months ahead, whether a customer is getting value and is likely to stay. It’s also what my team directly works to move — sending customers the right insights through the right channels about what they’re using and what they’re not yet using, so we’re actively closing that gap. If someone’s using only a fraction of what they’ve invested in, that’s a signal to act now. And if they’re deeply adopted, satisfaction and renewal tend to take care of themselves.
EITK: What technology trend are people underestimating right now?
Jyotsna: This one isn’t about a specific model or chip; it’s about where competitive advantage actually comes from now.
For a long time, the moat was your product or your data. But AI has changed that: product ideas can be copied and rebuilt faster than ever, and data alone is everywhere. So in my view, the real moat now is context, and what I’d call contextual velocity. Context is how deeply you understand your users; contextual velocity is how fast that understanding flows through your organization so you can act on it and turn it into relevant experiences for customers. When anyone can copy the product, what differentiates you is the context you hold and how quickly you can put it to work. People are only starting to realize this, but it’s becoming the real differentiator.
