Frank
I started at The Medical Center of Central Georgia in Macon when George W. Bush was still in his first term. The servers had physical tape drives. People printed things. The help desk ran on gut instinct and a shared spreadsheet nobody trusted.
That organization is now called Advocate Health. It’s one of the largest nonprofit health systems in the country. And I’m still here, same building, different job title, watching the wheel spin.
Most IT career advice is built around movement. Change jobs every two years, stack titles, chase the market. There’s a version of that wisdom that’s genuinely useful. But it ignores something important: what you learn when you stay.
I’ve been a Computer Tech, Systems Analyst and Systems Engineer here for nearly seventeen years, now specializing in Microsoft Exchange Hybrid and Active Directory, supporting an IT org of about 2,500 staff across an enterprise that covers 162,000-plus employees. That kind of scale gives you a specific kind of education. You don’t just see trends arrive. You watch them get implemented, hyped, defended, quietly abandoned, and sometimes resurrected under a new name. The full lifecycle. Most people who job-hop every couple years only catch one or two chapters of that story.
Here’s what three generations of IT trends actually look like from one fixed point.
Generation One: Virtualization and the Great Hardware Consolidation
When I arrived, physical servers were king. Every application wanted its own box. The server room was loud, hot, and full of machines doing one job badly. Then VMware showed up and IT departments everywhere acted like they’d discovered fire.
Virtualization was real. The consolidation benefits were real. But the way organizations implemented it often wasn’t. The tendency was to virtualize everything immediately, including things that had no business being virtualized, and then act surprised when performance tanked. We learned, eventually. But the learning cost something.
The pattern I started noticing back then: a technology arrives that genuinely solves a real problem, and the enterprise response is to immediately overextend it into problems it was never designed to solve. Every generation of IT trend follows this same arc. The tool is good. The deployment strategy is chaotic.
Generation Two: The Cloud Stampede
This one is still happening, which tells you something.
Around 2010 to 2015, the pressure to move everything to the cloud went from a conversation to a mandate. I watched organizations gut on-premises infrastructure because a vendor told them the cloud would be cheaper and simpler. Sometimes it was. Often it wasn’t.
What actually happened: hybrid environments became the default, not by design but by necessity. You can’t migrate a 162,000-employee health system to the cloud on a Tuesday and call it done. You end up with one foot on each side of the river, and someone has to keep that working. That someone, in a lot of organizations, is the Exchange administrator who actually understands how mail flow, hybrid connectors, and Active Directory sync interact when something breaks at 11pm.
I run Exchange Hybrid. Have for years. Every time someone declares Exchange is dead or that on-premises email is irrelevant, another enterprise somewhere hits a hybrid configuration problem that requires actual expertise to untangle. The cloud is real. Hybrid is also real. The idea that one replaces the other cleanly is a sales pitch, not a plan.
The Part Nobody Talks About When They Tell You to Chase Trends
Here’s the thing about watching trends from one place for seventeen years: you develop a reliable early warning system.
The warning signs are almost always the same, regardless of the technology:
- The vendor’s marketing collateral arrives before the documentation does
- Your organization forms a committee before anyone has touched the product
- The people championing it loudest have the least operational experience with it
- The migration timeline is built backwards from an executive presentation, not forwards from actual complexity
I’ve seen this pattern with virtualization, cloud migration, SD-WAN, zero trust, and now AI. Every wave brings the same sequence. Announcement, enthusiasm, over correction, hard reality, quiet re-calibration.
The difference between someone who’s been in one enterprise for seventeen years and someone who has worked at five companies in that same time is this: I’ve seen what happens after the press release. Most people only see the before.
That’s not a knock on job mobility. Seeing different environments teaches you different things. But there’s a specific kind of institutional knowledge that you only build by staying through the consequences.
When we moved our Exchange environment into a hybrid configuration, it wasn’t a six-month project you finish and walk away from. It’s a living system that requires continuous attention, especially when Microsoft changes something upstream and the hybrid connectors start misbehaving in ways that would be completely opaque if you didn’t have years of context on how the environment was built.
That context doesn’t transfer in an offboarding document. It lives in the person who was there for the decisions, the workarounds, and the outages.
Generation Three: AI and the Familiar Feeling
I’m watching the AI wave now with the same eyes I watched the cloud wave with in 2012. The technology is genuinely impressive. Some of it is genuinely useful. The hype layer on top of it is completely disconnected from operational reality.
I use AI. Every day, actually. I’ve built real applications with Claude and Gemini APIs. HookHouse-Pro runs on multiple AI integrations and it does things I couldn’t have built otherwise. I’m not skeptical of the technology itself.
I’m skeptical of the rollout pattern. The enterprise AI strategy memos that arrive before anyone has a working prototype. The job descriptions demanding five years of experience with tools that are eighteen months old. The assumption that deploying a chatbot means you’ve addressed a workflow problem, when usually you’ve just put a friendly face on a process that was broken before the chatbot arrived.
The organizations that are going to come out ahead on AI are not the ones that moved fastest. They’re the ones that took time to understand what problem they were actually solving, tested it against operational reality, and built institutional knowledge around the implementation instead of just spinning up a subscription and calling it transformation.
I’ve seen what happens when you skip those steps. Twice.
What Seventeen Years Actually Buys You
It’s not comfort. I’ll tell you that clearly. Staying in one place doesn’t mean coasting. The scope of what I manage today compared to what I managed when I started is completely different. The tools are different. The scale is different. Healthcare IT in 2025 bears almost no resemblance to healthcare IT in 2008.
What it buys you is pattern recognition that doesn’t come from reading industry blogs or watching conference keynotes. It comes from living through the full arc of things enough times that the next arc starts to look familiar before the keynote is even over.
I work best under deadlines I pretended didn’t exist until 20 minutes ago. But on the big picture stuff, the trends and the technologies and the organizational decisions, I’ve learned to slow down and ask the same questions every time: what problem does this actually solve, who is going to maintain this when the vendor enthusiasm wears off, and what does the failure mode look like?
I’ve watched enough good technology get deployed badly to know that those questions matter more than the technology itself.
Seventeen years is a long time. Long enough to watch virtualization go from revolutionary to assumed infrastructure. Long enough to watch the cloud go from a bold bet to a complicated hybrid reality nobody totally planned for. Long enough to see the same mistakes made with fresh marketing copy.
The window I’m sitting at hasn’t moved. But the view keeps changing, and that’s the whole education right there.
