I didn’t start with a strategy. I started because I was drowning.
Eight months into a new role, managing a book of business across dozens of accounts, sitting in back-to-back meetings where every conversation referenced three previous conversations I was supposed to remember. I was getting good at the work. I was also getting buried by the volume of it.
The first thing I built wasn’t impressive. It was an anonymizer. I was feeding meeting transcripts into AI to help me find patterns across accounts — who said what three months ago, how it connects to what someone said today — but I didn’t want client data going anywhere it shouldn’t. So I wrote something that stripped names, dates, IPs, and replaced them with placeholders. Client becomes AURORA. Contact becomes Contact-1. The substance stays. The identifiers don’t.
It worked. I started getting answers in fifteen minutes that used to take me an hour of digging through emails and CRM records. People started asking how I knew things. I said I was magic. They laughed. I kept building.
The Part Nobody Talks About
Here’s what the AI evangelists leave out: the first six months are ugly.
You’re Googling error messages. You’re building something that works on Tuesday and breaks on Wednesday because of a software update. You’re explaining to IT why you need read access to your own calendar and watching them look at you like you just asked for the launch codes.
You’re also second-guessing yourself constantly. Is this worth the time? Am I overcomplicating this? Could I just use a spreadsheet? The answer to that last one is always yes — and it’s always the wrong answer.
The gap between “I could do this manually” and “I built something that does it for me” isn’t efficiency. It’s headspace. When the machine handles the recall, the pattern-matching, the prep work, your brain gets to do what it’s actually good at: reading the room, making judgment calls, connecting things that don’t obviously connect.
That’s the part that compounds. Not the automation. The cognitive space.
What Actually Changed
I’ll tell you the moment I knew this was real.
I was about to walk into a meeting with a senior leader. My AI assistant had prepped a summary — account history, recent conversations, open items, the stuff you’d want to know walking in. But I’d been deep in another project and hadn’t looked at it. I walked in cold. Unprepared. The meeting went fine because I’m experienced enough to wing it. But I knew I’d left value on the table.
That’s when I stopped thinking about AI as a tool and started thinking about it as a spotter. Not something that does the work for me — something that catches the ball I already threw but forgot about. Something that says: you’ve got fifteen minutes before your next meeting, here’s what you prepped last week, here’s what changed since then.
I didn’t need AI to be smarter. I needed it to make sure my own intelligence showed up on time.
The Political Part
Nobody gives you permission to do this. You just start.
You build quietly. You anonymize everything. You stay inside the lines. You don’t pitch it as a revolution — you pitch it as “I just want my calendar to remind me about meetings.” When people ask how you knew something, you don’t give a TED talk. You say it’s working and move on.
Then something shifts. Someone in leadership notices you’re producing work faster than people who’ve been there three years. Someone in IT notices you’re asking thoughtful questions about permissions instead of just brute-forcing access. Someone pulls you aside and says: what are you doing? Can we do that for everyone?
That’s the moment you’ve been building toward. Not because you planned it — because you were solving your own problem loudly enough that other people recognized theirs.
What I’d Tell Someone Starting Today
Don’t start with the technology. Start with the thing that’s drowning you.
For me it was account context — holding the state of twenty-plus client relationships in my head simultaneously. For someone else it might be email triage, or meeting notes, or finding the document they know exists but can’t locate. The specific pain doesn’t matter. What matters is that you pick the one that costs you the most cognitive energy and you solve it first.
Then you solve the next one. And the next one. And six months in, you realize you haven’t built an AI system — you’ve rebuilt how you work. The AI is just the scaffolding.
The people who are ahead right now aren’t ahead because they’re smarter. They’re ahead because they started. They were willing to feel stupid for a few months while they figured out what the tool actually does versus what the marketing says it does. They built ugly things that broke. They fixed them. They kept going.
The gap is going to widen. Not because the technology is hard to learn — it isn’t. Because the people who started six months ago have six months of compounding returns on their investment. Their AI knows their accounts. Their workflows are dialed. Their institutional knowledge is indexed and searchable. That advantage doesn’t come from a product demo. It comes from showing up every day and feeding the machine one more piece of context.
The Thing I Didn’t Expect
The biggest change wasn’t professional. It was personal.
When you stop spending all your energy just trying to keep up, something else has room to surface. You start noticing things. Patterns in your own behavior. Ways you’ve been compensating for years without realizing it. The AI becomes a mirror — not because it’s profound, but because it’s consistent. It reflects your thinking back at you without the noise.
I started building tools to help me work. I ended up building tools that helped me think. The line between those two things turned out to be a lot thinner than I expected.
I’m not writing this because I’ve figured it all out. I’m writing it because I’m eight months in and the trajectory is unmistakable. The people around me are starting to see it. The organization is starting to move. And the thing that started as a calendar reminder is becoming something I couldn’t have imagined when I typed that first prompt.
Start with the thing that’s drowning you. The rest follows.