AI has become a pressure point for every Infrastructure & Operations (I&O) leader. CEOs are being told AI will define competitive advantage. CIOs are being pushed to modernize faster than budgets, teams or architectures can realistically support. And almost every part of the business now expects I&O to “enable AI” even as day-to-day operational demands continue to rise.
In a recent Netskope survey carried out across I&O leaders, 80% said their organization’s infrastructure is now central to delivering core business goals. Almost half are expected to support AI-driven innovation. Yet only 38% feel fully equipped to meet those expectations, with most operating reactively while trying to keep the lights on.
Two distinct priorities – delivering perfect reliability and accelerating innovation – are a persistent tension for I&O teams. And most are dealing with foundational issues like scattered knowledge and documentation, inconsistent device visibility and the operational drag created by hybrid work.
AI Won’t Wait for Your Infrastructure to Be “Ready”
Under normal circumstances, these gaps might justify delaying AI initiatives. But delaying is no longer an option. AI experimentation is already happening in every organization, whether there’s a strategy or not. If I&O doesn’t guide it, employees will experiment on their own, often without guardrails.
So how do you move the organization forward?
At Netskope, our answer was simple – create a safe space to experiment.
We set out to rapidly scale safe, meaningful AI adoption across the organization by giving every part of the business a structured, governance-approved space to experiment, learn and build real AI-driven improvements.
A core part of my role as I see it is to make it easy to work at Netskope. To this end, we wanted to set the tone for AI adoption in every area of the business to give teams outside of engineering the opportunity to get hands-on with the technology and to get excited by it.
Importantly, we wanted to make a lasting impact. This was about creating deployable AI accelerators for the business and the people within it.
Instead of waiting until every process or governance model was perfect, we created the right boundaries, the right protections and a controlled environment where people could learn without putting data at risk.
Inside Netskope’s Promptathon
When I say Promptathon, think about a hackathon that is dedicated purely to AI – a collaborative event where people come together to experiment with and refine AI prompts.
The principle behind it is straightforward. If you want employees to adopt AI responsibly, you have to give them the chance to try it.
So, we invited the entire organization to come together to practice prompt engineering. Participants worked in small teams to create and test prompts that addressed real business challenges they had identified themselves, and then they iterated quickly and collaboratively to get to the best results.
Finance, HR, procurement and sales all took part. Functions that rarely experiment with emerging technology suddenly had a way in.
In fact, nearly 600 people from 17 departments across 20 countries took part. And we had nearly 100 project submissions ranging from AI projects that helped sales teams personalize pitches; accelerated quality assurance processes; and auto-generated documentation from recorded videos. Everyone who joined in got a T-shirt, which encouraged people to participate and helped to create a brand around the event.
The outcome? We changed AI adoption in our business overnight.
AI usage across Netskope doubled and stayed there. Adoption didn’t spike and drop. It permanently reset our baseline.
Employees kept using AI because they had seen its value first-hand, learned from each other, and built solutions that addressed real problems in their workflow.
And we are building a post-event library to provide access to all the reusable artefacts from the submissions.
Experimentation Requires Trade-Offs
I know what you are thinking. That sounds great, if you have time, but we are too busy to indulge in such soft innovation adoption tactics. Perhaps you are musing that it’s all very well and good for a tech company, but it’s harder to justify the time in finance, healthcare and retail. Competing priorities aren’t new for I&O, but the actual trade-offs involved rarely get talked about.
I will be completely honest; running the Promptathon meant temporarily deprioritizing some operational work. Not because those tasks weren’t important, but because enabling the organization’s future mattered just as much. We knew that the window of opportunity to accelerate the use of AI across the organization was closing.
This is where simple, risk-based reasoning becomes essential. When I assess whether we can pause something, I start with one question: What’s the worst thing that happens if we wait?
If delaying a laptop shipment blocks a new hire entirely, that’s unacceptable. But if we can mitigate the delay with a temporary virtual desktop, the worst case becomes tolerable. That gives us room to redirect capacity.
This is the reality for I&O today. You can’t do everything at once. You need to understand impact, risk and alternatives and make conscious choices instead of reacting to noise.
The Promptathon was a perfect example. It stole time from ‘business as usual’, but we were able to make the case – in terms that the business understood – that the long-term productivity gains far outweighed the short-term slowdown. We made that call deliberately, with executive support, and the sustained jump in AI use in all areas of the business more than justified it.
What We Learned: The Real Levers for Driving AI Adoption
Five factors made the Promptathon both a success on the day but also in the long-term, and they are easily replicable.
1. Executive sponsorship removes hesitation
Our CIO led the internal communications and stayed involved throughout. That signal alone cut through uncertainty and created immediate momentum. It was clear to everyone that this was no side project.
2. Tangible incentives matter more than you may think
Never underestimate the power of a T-shirt. Small, visible tokens of belonging can dramatically increase engagement in AI initiatives, particularly for non-technical teams who may feel uncertain or intimidated.
3. Safe enablement beats restriction
We used an enterprise-grade AI platform where employees didn’t have to worry about data risk. When people feel safe, they experiment. When they experiment, they learn.
4. Recognition fuels momentum
The prize wasn’t the draw – visibility was. Leadership awareness and giving the seven winning teams the chance to showcase their projects in TED-style talks helps embed cultural adoption and creates a cascade of new ideas and use.
5. Plan how you will maintain momentum
A Promptathon shouldn’t produce one-off demos. Create a mechanism to store, refine and operationalize the results so the value compounds. Momentum dies immediately if you treat the event as the end point. And remember, the biggest AI productivity gains often sit outside engineering. You need to lower the barrier to experimentation for “business-side” functions.
Underneath all of this is something simple – adults learn by doing. If you want to accelerate adoption, create the conditions for experimentation and get out of the way.
Lead by Enabling, Not Controlling
For I&O teams facing the balancing act of performance and innovation every day, structured experimentation is the way forward.
By giving employees a safe place to test new tools, you reduce shadow IT, uncover genuine use cases and avoid the paralysis that comes from waiting for readiness that may never come.
And critically, it positions I&O exactly where it should be in the age of AI – at the centre of progress, not chasing after it.
About The Author Of This Article
Elena Matchey is Senior Director of IT Infrastructure & Operations at Netskope
Also Read: The End Of Serendipity: What Happens When AI Predicts Every Choice?
[To share your insights with us, please write to psen@itechseries.com ]
Comments are closed.