Last month, I made a hiring decision that would have been unthinkable three years ago. Not because it involved AI, though it did, but because I made it while sitting on a chairlift in Whistler.
I was holding my snowboarding gloves in one hand, and my phone in the other, interrogating a financial forecast before approving the budget for a new hire. I was pressure-testing a machine learning-generated utilization prediction, challenging its confidence intervals, and cross-referencing cash flow scenarios generated by an orchestration layer I helped build.
The budget decision took minutes. The old way of decision-making—using spreadsheets, gut instinct, and waiting for month-end reports—would have taken weeks and probably would have been wrong.
Let me put this in context. I’m the CEO of Transcends, a 30-person creative intelligence agency, and for the past several months I’ve been deep into developing something I call our autonomous finance plan. It’s an integrated stack of SaaS tools including QuickBooks, Jirav, Scoro, and Dext, wired together with AI and machine learning to replace the Frankenstein mess of spreadsheets and disconnected tools that were strangling our ability to make good, fast decisions about budgets for hiring.
But here’s the thing that’s been rattling around in my head since attending Microsoft Build in June. It wasn’t the shiny new agent demos or the technical announcements. It was one line from Satya Nadella, Microsoft’s chairman and CEO, that felt like he’d been reading my mind: “AI alone won’t change your business. The system running it will.” My agency didn’t just need AI tools; it needed the system AI would operate within.
Tech isn’t the bottleneck
Microsoft announced at Build that its Agent Framework has reached general availability and multi-agent orchestration is production-ready in Copilot Studio. Cue the breathless headlines.
But here’s what the tech press won’t tell you: you can buy all that tomorrow, and it still won’t matter if your CEO or leadership team isn’t willing to be engaged in technical day-to-day work. Agentic AI can’t just be theoretical; it needs to be experienced and made human.
The “build it or buy it” question matters strategically. The underlying machine learning models, whether they be gradient-boosted trees, time-series forecasting, or anomaly detection, are increasingly commoditized. Any competitor can license the same stack.
The real bottleneck in AI adoption isn’t technical. It’s organizational. It’s human.
It’s whether the CEO and leadership team are willing to roll up their sleeves. Not just approve the AI strategy from a slide deck at the quarterly offsite, but participate in building workflows, interrogating forecasts, understanding what models can and cannot do, and making peace with probabilistic thinking instead of the false comfort of deterministic spreadsheets.
A recent quote from Dr. George Westerman, senior lecturer at the MIT Sloan School of Managementnailed it: “If you think about digital transformation as two words, we pay too much attention to the digital and not enough to transformation. It’s not a technology challenge, it’s a leadership one.” When I talk to fellow founders and executives, they’re excited about AI in theory. But when I ask if they’ve actually used the tools they’ve purchased, trained a model, or debugged a prediction, it’s crickets.
Orchestration, not ownership, wins
Microsoft’s approach at Build made this concrete. They’re not just shipping models—they’re building what Jay Parikh, EVP of CoreAI, calls “a comprehensive agent platform” that spans GitHub, Azure, Microsoft IQ, and Foundry as a single integrated system.
That’s the insight. The magic isn’t in any one piece; it’s in the orchestration layer of how agents are built, contextualized with enterprise data, governed in production, and improved continuously under human oversight. This is exactly what we’re doing at Transcends with our autonomous finance plan, just at a different scale.
It isn’t proprietary AI. We’re retaining core SaaS platforms and building our own orchestration on top. The predictive models inside those platforms are good old machine learning, not frontier models. But the system is what gives us an execution advantage.
And I do mean execution advantage, not a permanent moat that enables a company to protect its market share. Our autonomous financial plan doesn’t give Transcends monopoly power, but it keeps us learning, adapting, and in the race as an agency inspiring creative intelligence.
AI fluency is essential
Here’s the truth as I see it: in 2026, your willingness to be hands-on with AI is becoming as important as your ability to read a P&L.
Research on AI and competitive advantage increasingly suggests that as AI capabilities become broadly accessible, competitive advantage shifts toward complementary organizational assets, including workflow redesign, governance mechanisms, managerial capabilities, and leadership’s ability to integrate AI into decision-making and operating models.
The companies winning aren’t those with the fanciest demos. They’re the ones whose leaders understand systems thinking, can translate between technology and operations, and are comfortable making decisions under uncertainty rather than waiting for perfect information. This is why the skill of entrepreneurship is so valuable today.
At Transcends, our financial plan has already delivered tangible returns: we’ve avoided hiring an additional full-time finance employee, we’ve saved roughly 12 hours per month of our finance manager’s time for strategic work and caught early warning signals that would have been invisible in spreadsheets. But the deeper shift is cultural.
Build, buy, or bail?
The AI transformation every consultant is selling you isn’t a technology challenge. It’s a management challenge. A leadership behavior challenge.
You can buy the Microsoft Agent Framework tomorrow. You can license the same SaaS stack we use. You can hire the consultants and approve the budget. But if you’re not willing to be in the system testing, breaking, learning, and iterating, you’re not adopting AI. You’re just rearranging furniture.
The question Microsoft Build left me with isn’t: What should I buy? It’s this: When AI reshapes how decisions get made in your business, will you be the one making them, or will you be the last to know?
Ashleigh Vogstad is CEO of Transcends.