Every business has one. The person who knows how everything actually works. The one everyone goes to when something breaks, when a process doesn't make sense, when the documentation doesn't match reality.
They're invaluable. They're also a massive risk.
Not because they'll do something wrong—because they can't be in two places at once. They can't take a proper holiday. They can't get sick without everything slowing down. And eventually, they'll leave, retire, or burn out. When that happens, years of institutional knowledge walks out the door with them.
How they got here
This person didn't set out to become a bottleneck. They became one because they cared. They stayed late, learned the systems, built relationships, solved problems nobody else wanted to touch.
That's the uncomfortable part. The same qualities that made them invaluable created the dependency. Rewarding them with more responsibility just made it worse.
They know the history—why that process exists, what happens when you skip a step, which workarounds matter and which ones are just habit. They're the living documentation. And they're exhausted, spending most of their time answering the same questions, explaining the same context, catching the same mistakes before they become problems.
Elevation, not replacement
AI isn't about replacing this person. It's about capturing what they know so they can stop being the bottleneck and start being the one who makes the system smarter.
Most of their routine work—the repeated explanations, the error-catching, the translations between systems that don't talk to each other—can be captured well enough to handle the common cases. Not through documentation, which is static and nobody reads anyway. Through systems people actually use: an AI assistant that handles the routine questions, with the expert reviewing and correcting until it gets things right on its own. Their pattern recognition turned into dashboards the whole team can see.
Once that knowledge is embedded in how the business operates—not locked in their head, not buried in a wiki nobody checks—it compounds. Every correction makes the system sharper. Every captured explanation is one less interruption.
They get a better job
It's not just about de-risking the business. It's about giving this person their time back.
The hard problems. The judgment calls. The stuff they never had time for because they were explaining how the invoice process works for the hundredth time. Maybe they've been stuck in the same domain for years because the business couldn't afford to move them. Now they can. Instead of being buried in internal operations, their expertise becomes visible—the person clients want in the room, not because they're the only one who knows, but because they're the best.
You probably thought of someone while reading this. That's the person whose knowledge is at risk, and whose potential is being wasted on repeat work. If you want to talk about what capturing that knowledge looks like in practice, we're here.