A New Type of Real Estate Website — Built by an Agent, Powered by AI, Written for Consumers.

I’m a real estate agent. I have no technical background, no development team, and no business building a website from scratch.

But I had a vision — and it started with a frustration that had been building for years. Every time I sit down with a new buyer or seller, we spend the whole first meeting on basics. Not because my clients aren’t informed — they’re some of the most educated people in the country — but because our industry does a terrible job of explaining itself. The information that exists online is either national content that doesn’t apply locally, or local content that lacks practical information you can actually use. Data without interpretation. Process without context.

I wanted to build something different: a place where a buyer or seller in the DC metro could spend time at their convenience, before we meet, and get a clear idea of what they’re getting into. How costs compare between DC, Maryland, and Virginia. What to expect from an agent. What’s negotiable — and what you might be giving up when you negotiate. How to evaluate neighborhoods and schools on your own terms. A resource that should have existed already but somehow still didn’t.

I hired a developer. It took weeks to find the right one — I interviewed many, even started working with a few, but had to stop within days when it became clear they weren’t the right fit. When I finally found someone who seemed to understand my concept, it felt like I was close — close to finally having the website I’d been thinking about for years.

But after months of going back and forth, a pattern emerged: we were speaking the same language, but the end result in his mind and mine kept looking completely different. He delivered a visually stunning site that wasn’t built to hold the substance I needed. Meanwhile, I’d been quietly watching the backend — at first out of curiosity, then because I had a gut feeling I might need it.

It was clear we were going in circles. Time to make a decision. I mean, how hard can it be? Although I was slightly scared of what I am committing myself to, I refused to accept that this website would be yet another version of “well, this is what works why change it? websites don’t matter anyways.” I took the site he delivered and set out to tweak it. Add some text here, delete some boxes there, not a piece of cake, but doable if I put in the effort. And I can put in the effort when I’m interested — thank you, neurodivergence, which at times makes you a superhero and at other times makes you question your life choices all at the same time.

But with every edit and every page, the beautiful site became at first a little asymmetrical, then downright amateurish. After yet another 12-hour day, I looked over what I’d achieved and wanted to cry and call it quits. I’ll ask the developer to restore it, I thought. He’ll roll his eyes, but he’ll do it.

Defeated and exhausted, I was turning off my computer when I saw Claude’s app still open.

Until that moment, my use of AI — Claude, ChatGPT, Gemini — had been limited to writing and researching. But I’d kept reading about Claude being very good at coding. Coding... coding... and an idea came to me. That section I’d been fighting with all day, where one line of text kept sitting misaligned for reasons I couldn’t figure out — what if I showed Claude what I was trying to do and it could code it correctly?

I remembered seeing a widget in my website builder called “HTML” — which I knew was a coding language. I went back to the builder, found the widget, then asked Claude to explain what the widget does and how to use it. That felt promising. Then I asked if we could code this one section and add it to the page. We could. I copied the code, pasted it into the widget, and went silent for a moment.

It was after 1 AM. My house was quiet except for my loyal pug snoring up a storm at my feet. I was exhausted — emotionally, physically — but I knew I wasn’t done. I wasn’t going back to the developer. And I might just pull this off and create the site I envisioned.

Weeks went by, then months. To my husband I seemed a bit possessed. My daughter was curious but mostly unclear on what I was doing. To be honest, I wasn’t entirely clear either.

What I can say is that I was learning — though not what I expected. Was I learning to code? Not really; Claude was doing that. But I was learning the process of building a website: the power of Google’s SEO algorithm, what makes a site technically sound versus what makes it good for the person using it, how to communicate a vision to AI clearly enough to get it right, and how to challenge every output rather than trust it blindly.

More importantly, I was learning how to use AI to do something so technical that I couldn’t explain it to my friends or my husband. How can you, when you don’t fully understand what’s happening yourself? So, despite making continuous progress, to everyone around me it wasn’t clear what the heck I was doing or why I needed to pull 12-hour days when nobody was making me do it and I wasn’t on any deadline.

But I felt like I was on fire. And here’s the other thing about being neurodivergent — you want perfection when it doesn’t exist, and you can’t stop until you get it, or learn the hard way that ‘very good’ is as perfect as it gets.

What started as small coded sections turned into entire pages — and each rebuild captured where I was in the process of learning what AI and I could do together. The first version was a monstrosity: the leftover developer site stitched together with coded sections completely out of sync. The second was the widget era — I built a ton of them, just because I could. That’s when my husband gave me the kind of honest feedback only a spouse can deliver: the widgets are impressive and some of them, like the closing cost calculators, are very useful. The rest are fun but don’t add value. “You have genuinely useful and interesting perspectives, but way too many details that are hard to skim through. I as a consumer don’t really have the time for that — I want to understand the general process. And the depth, the sourcing, the transparency — that already tells me I can trust you and you know what you are doing.”

He was right, of course. Each version after that got leaner — trimming text, replacing dense explanations with infographics, refining design and user experience. And AI kept getting better with each update, capable of more than I could even understand. The only real limitation was in me — finding the right way to explain what I wanted and keeping a constant eye on hallucinations.

And somewhere in that process, something shifted. I went from being an observer of somebody else’s work to being the one making the decisions about what works and what doesn’t. Not because I became a developer — I didn’t. But because AI made it possible to cherry-pick: learn what you need to understand, delegate what you don’t need to do yourself, and still produce something professional. That’s not how learning has ever worked before. It’s not how building has ever worked before. And it’s available to anyone who wants it.

So why does any of this matter beyond my own story?

Because the reason a site like this doesn’t yet exist isn’t the novelty of the idea. Plenty of agents wish they could offer something like it. The problem has always been that it requires two professionals in one: a real estate agent who understands the local market deeply enough to create genuinely useful content, and a web designer who can understand real estate enough to build it. Those two skillsets almost never overlap. Most individual agents or small teams don’t have the capital to invest in it, bigger teams don’t need to, and most brokerages use their websites for marketing and positioning, not real-life education. AI bridged the gap — and the bet I’m making is that current consumers want and need that information and that both the real estate industry and our clients will be better off if transparency like this exists.

And this is exactly why I created this website: every time I work with a buyer or seller, I have the same thought: consumers don’t understand our industry. Even people who genuinely follow real estate have a skewed view of it, shaped by reality TV shows or sixty-second social media takes that can only scratch the surface. And if we’re really honest, the industry itself is not always on the same page — not everything makes sense and a lot of it needs changing to adapt to modern realities.

So the idea I am betting on is that we must have a single place where a potential client could get a factual, thorough picture of what buying or selling a home in the DC metro (or any other market for that matter) entails — and come to our first meeting ready to talk strategy instead of basics. Because if a client reads all of that before we sit down, it saves us both time, prevents misunderstandings, and the client knows, not guesses, that they are getting value. Real estate isn’t rocket science, but it has a lot of moving parts, a lot of possible strategies that yield different outcomes, and a lot of outside influences that neither an agent nor a seller/buyer can predict. Knowing about those from the start is the only way to a successful partnership.

And that’s how we got here, 11 months later.

Is this website in its last version? Definitely not. Will people find mistakes? I am sure — although I can honestly say I have not cut any corners in verifying each fact and made sure to cite the sources I used. Is this a useful resource for my clients? I very much hope so.

So what’s next? Putting it out there and finding out if I’m right. Are consumers looking for this kind of resource — the knowledge to make better choices with their biggest investment? Or am I misreading the data and drawing conclusions that fit my own worldview? Could be. But I’d rather put it out there and find out than spend another year wondering.

But for now, I’m proud of what I created. I gave it all.