AI is everywhere in theory, but in practice, it’s still stuck in trial mode for most companies. A handful of pilots here, a few licenses there, maybe some clever email rewrites. But that’s not a strategy; it’s dabbling.
The numbers speak for themselves. A new MIT study found that 95% of enterprise GenAI pilots fail to deliver measurable business impact. In other words, most companies are running experiments that never scale.
AI can absolutely make individuals more productive. But the bigger prize is changing how the entire organization operates. The challenge is that most leaders lack the knowledge to scale from interesting experiments to genuine enterprise transformation.
That’s where the Center of Excellence (COE) comes in. At first, the name might sound like centralization for its own sake. But done right, a COE isn’t about control; it’s about inspiring and enabling. It’s a model that gives every team the confidence, tools, and freedom to reinvent how they work.
Start with the right people
The first mistake companies make when building a COE is staffing it with only IT leaders. AI is not just a technical challenge; it’s a business challenge. That’s why the people at the heart of your COE need to be your natural disrupters—the employees already tinkering with AI in finance, legal, customer service, marketing, and operations. These are the people who are constantly asking themselves if there is a better way to get things done.
At Alteryx, we built our COE by starting small but credible, pulling in cross-functional voices from finance to product to the C-suite. This mix ensured we weren’t just chasing the latest tools but tied everything back to real business value in a measurable way.
Enable, don’t control
Too many Centers of Excellence become gatekeepers, slowing down the process and creating frustration. A true AI COE is an enabler. Its job is to spark curiosity, share best practices, and show teams how to build their own solutions rather than issue tickets or approvals. It’s about enabling the folks who know the business, know the data, and know the process to super-charge or even completely re-think the process with AI.
One of my favorite examples of a solution built through the COE at Alteryx is a workflow we created with the go-to-market team. It pulls data from our CRM, marketing automation, and product telemetry. Before a customer meeting, our salespeople can ask this LLM what’s going on, and it will summarize renewal dates, deployment stats, and open support tickets in plain English.
Even better, we’ve taught employees how to build their own GPTs with the help of Alteryx Copilot. Suddenly, AI isn’t intimidating. It’s approachable, usable, and even fun.
When employees can see hands-on training in action, their mindset shifts. They stop asking if AI will take their job and start imagining how AI can make their job more valuable.
Look for the “aha” moments
Every function has a tipping point when AI finally clicks. For a finance team, it might be building faster and more accurate valuation models. For a legal department, automating paralegal-level tasks that would otherwise consume hours is most valuable. For a sales organization, smarter outreach and preparation before meetings are key.
The COE’s role is to help teams discover those breakthroughs and then scale them across the business. Because once people see what’s possible, adoption spreads quickly.
And to anyone worried about AI replacing their career, the reality is that none of us know exactly how roles will evolve. Your plan for the next 10 years shouldn’t be to look for a company that hasn’t heard of AI. The best long-term strategy is to become comfortable with scaling your work using AI. That is how you future proof your career. Lean into environments where you can figure out how your role evolves with AI, take advantage of any training provided, and get used to constantly learning.
From experiments to enterprise impact
Scattered pilots or a siloed AI team will never move the needle. Real transformation happens when the COE shifts from gatekeeping to enabling. That means equipping employees with the right tools and training, setting smart guardrails, and keeping leaders engaged and accountable.
When that happens, AI stops being an experiment and becomes the way a company actually operates. And that’s when customers start to feel the difference.
AI will not transform your business from the sidelines. It will transform your business when curiosity, experimentation, and empowerment become cultural norms. That’s what an AI Center of Excellence is designed to create: a safe environment where employees learn, adapt, and grow.
The sooner you build a COE, the sooner AI stops being “the future” and starts becoming how your company works today.
Andy MacMillan is CEO at Alteryx.