This past week I did something different.
Instead of talking about what AI can do for business owners, I just showed them.
Friends. Old coworkers. Real estate operators at a networking event. Trades business owners at an industry event. People who understood AI conceptually, had maybe played with ChatGPT a few times, but had never seen what it looks like when someone actually sits down and builds with it.
The reactions told me everything about where we are.
Then this morning Anthropic announced Claude Managed Agents. And the bar dropped even further. No, it’s not the Mythos model that people think will destroy the world.
What I showed them
Over the past several months I have built a lot of things.
PassTheKeys.ai — a platform for capturing and transferring business owner knowledge before they exit. LandOps — an agentic CRM built specifically for land and real estate operators. StormOS — an AI system that tells restoration companies which homes are most likely to need their services before the homeowner even knows there is a problem. A mobile ecommerce app. Personal assistant agents and operating systems for business owners.
Different industries. Different use cases. Different problems being solved.
One thing in common: Claude built them with me.
No development team. No agency. No six-figure software budget. Just domain expertise, knowing the problem well enough to describe it clearly and the tools that exist right now to bring it to life.
When I put these in front of people this week, the question I kept hearing was not "how long did that take" or "what did it cost." It was "you built that in Claude?"
Yes. That is the point.
The moment that stood out
I was at a real estate networking event this week. Casual setting. I pulled up LandOps on my laptop — the agentic CRM I built for land investing and deal management and started walking a few people through it.
One guy leaned in about two minutes in and asked a simple question.
"Does it have to be real estate?"
No. Same build, different domain. You tell it how your business works instead of how a land deal works.
He went quiet for a second. Then he started talking about his business. Not asking about AI. Not asking about tools or costs or timelines. Just talking about his business, the problems, the gaps, the thing he had always wished existed but assumed someone else would have to build for him.
That is what showing beats telling. He was not convinced by an explanation. He was convinced by seeing something real and recognizing his own problem in it.
The one that stopped a trades owner cold
At a separate trades event I described StormOS.
The concept is simple but the implications are not. StormOS pulls public property data, layers in weather triggers, and runs a risk scoring model to identify which homes in a given area are most likely to need restoration services from roof damage, water intrusion, mold risk and delivers warm leads to a restoration company within hours of a storm hitting.
The homeowner does not know there is a problem yet. The restoration company already has their address.
The reaction was immediate. "What other niches can you build this for?"
That question is the AI operating model I help bring to life. The same architecture — public data plus event triggers plus risk scoring plus automated outreach. It works for any trades business that benefits from reaching a customer before they start searching. Roofing. HVAC. Remediation. Pest control. Tree services.
The technology is not the insight. The insight is understanding your customer's problem well enough to know when it is about to happen. That is domain expertise. That is the ingredient that makes the system work. The build follows from the knowledge, not the other way around.
What Anthropic just announced
Yesterday, Anthropic launched Claude Managed Agents. And it matters for every business owner who has been watching this space and waiting for the right moment to build something.
Until today, deploying an AI agent in production required months of engineering work that had nothing to do with what the agent actually did. Infrastructure. Sandboxing. State management. Credential handling. Error recovery. All of it had to be built from scratch before a user ever touched the thing. That engineering overhead is also what kept an entire layer of AI wrapper software in business — tools that existed specifically to paper over that complexity so non-developers could get something working.
Managed Agents eliminates most of that. You define what you want the agent to do, what tools it has access to, and what guardrails it should follow. Anthropic runs it on their infrastructure. The wrapper software layer just became largely redundant. So long Zapier, Make and n8n? Only time will tell. The feature releases are daily at this point and even someone who is in the trenches daily on ai news, it’s getting overwhelming, but exciting.
For the business owners I work with that means this: the personal assistant agent, the custom lead generation system, the AI operating model built around how your specific business works — all of it can now be deployed and run without a development team managing the plumbing underneath it.
The bottleneck just got smaller again.
What this means for you
The question used to be whether something could be built. That question is gone.
Then it was whether you needed a developer to build it. That question is almost gone.
Now the only real question is whether you know your business well enough to describe the problem clearly. Because if you do, the tools to bring it to life exist today, they are accessible, and the infrastructure to run it at scale launched this morning.
No code required. Just what you already know about your business.
If you want to see this firsthand I am hosting a free live webinar on Wednesday April 22nd at 11am Central called Build Your First AI Assistant in 60 Minutes. We will build something real, live, in the session. If you have an idea you have been sitting on, this is the week to do something about it.
See you next week. Enjoy Masters weekend.
Nick
Built, Not Bought is a weekly newsletter about building AI systems that run your business. Subscribe on LinkedIn or Beehiiv so you don't miss an issue.
Tired of news that feels like noise?
Every day, 4.5 million readers turn to 1440 for their factual news fix. We sift through 100+ sources to bring you a complete summary of politics, global events, business, and culture — all in a brief 5-minute email. No spin. No slant. Just clarity.