An agent that finds local businesses with no website, researches each one, and builds them a draft site — so a web designer can call and say "I already made you something."
Two videos · Built with Claude Code · ≈ 10 min each after the cut
Work in progress — parked here so it doesn't get lost.
What if you could build an AI agent that finds businesses that need websites, builds the website for them, and then gives you the owner's contact information so you can offer it to them?
That's exactly what I'm going to show you how to build today.
And we're not talking about an automation that just follows the same steps every time.
We're building an AI agent.
You give it a goal — like finding local businesses in a particular industry that don't have websites — and the agent figures out how to complete the job.
It searches for the businesses.
It identifies which ones don't have websites.
It gathers information about the business.
It builds a website specifically for that business.
And then it gives you the information you need to contact the owner.
And there's a real market behind this.
More than 6 million small businesses in America don't have a website.
You don't need 6 million. You need a few.
But here's why I really want you to learn how to do this.
The website agent is just one example.
Once you understand how to build an AI agent that can reason through a goal, use tools, and take actions to accomplish it, you can start thinking about agents completely differently.
An agent that researches potential customers.
An agent that qualifies leads.
An agent that creates content.
An agent that analyzes competitors.
An agent that handles parts of your customer support.
The opportunity isn't this one agent. It's learning how to build agents.
Because once you understand the process, you can start asking a completely different question:
What else could I build an agent to do?
And that's what I want you thinking about throughout this video.
I'm going to walk you through this entire build step-by-step, and you don't need to know how to code.
By the end, you'll have built your first working AI agent — and, more importantly, you'll understand the foundation you can use to start building your next one.
Let's build it.
"More than 6 million" is defensible — current data puts roughly 27% of US small businesses without a website, which is closer to 9 million. But some sources say 17%, which lands just under 6 million.
Safer and stronger: "Roughly one in four small businesses in America still doesn't have a website." Better supported, and a fraction lands harder than a number that big.
What viewers follow along with. This is the URL you say out loud.
drerinjacques.com/site-scoutEvery step, every prompt, no filming notes. Open it →
package-lock.json in your home folder — it makes Turbopack print a yellow workspace warning on every build. And the Vercel plugin will flag lib/agent.ts as an error when it is only a suggestion; turn it off or be ready to say so.
~/Desktop/site-scout-dryrun/output/ — the orthodontist site the agent built from four facts and no reviews. Open it from your laptop; don't host it, it's a real practice's name and details.