AI builds, then interviews: A conversation with Steve Wiese, real estate appraiser and inventor behind HomeValueOptimizer.com.
Steve Wiese, a seasoned real estate appraiser with more than 35 years of experience and multiple patents in valuation and real estate technology, recently launched HomeValueOptimizer.com — an AI-enhanced evolution of his earlier project, AVMOptimizer.com. In this interview, Steve shares how he used ChatGPT to bring his idea to life, what he learned from the process, and how AI is reshaping the future of property valuation.
Q1. What inspired you to create Home Value Optimizer in the first place? Was there a specific problem with existing home value tools that pushed you to find a better solution?
A: I’ve always been interested in innovation within real estate. Over the years, I’ve applied for and been granted several patents in both the technology and real estate sectors — developing apps and business processes that improve valuation accuracy, environmental assessments, and even property depreciation analysis.
As a real estate appraiser, it was natural for me to focus on improving how real estate is valued. When Zillow introduced the Zestimate, it changed how people thought about home values — and the public became fascinated, even obsessed, with those numbers. But I also noticed a major flaw in that otherwise great idea: the blind spot around upgrades and condition. That gap between algorithmic estimates and actual property realities is what inspired me to build Home Value Optimizer.
Q2. How did ChatGPT come into play during the development? Can you describe how much of the project was built with AI versus your own coding and appraisal experience?
A: Home Value Optimizer actually started as an evolution of my earlier project, AVMOptimizer.com, which was built on WordPress using traditional themes and plugins. I took the code from that original version and simply pasted it into ChatGPT, then began asking it to edit, simplify, and improve different parts of the code.
It became a real back-and-forth collaboration. I’d test what ChatGPT produced, give it feedback on what worked or what needed adjusting, and it would refine the code accordingly. ChatGPT also offered suggestions I hadn’t considered — ideas that improved both functionality and the user experience.
Readers might find it interesting to compare AVM Optimizer and Home Value Optimizer, since the latter was created largely with AI. The goal for Home Value Optimizer was to be a streamlined version — cleaner, faster, and easier to use, while still delivering the same valuation accuracy.
Q3. What were the most surprising or impressive things ChatGPT was able to do for you while creating Home Value Optimizer?
A: A couple of things really stood out to me — and they’re both interesting and cost-effective. ChatGPT was able to create custom code that replaced the functions of entire plugins, themes, and even tracking software. That kind of capability can completely change how websites are built, potentially saving designers and developers significant money while giving them more control.
One drawback is that you still need some basic web-development understanding to guide the process, but the more you work with AI, the more you learn — and I suspect the AI itself learns how to assist you better over time.
Another surprise was how well ChatGPT handled responsive design and device scaling. It automatically corrected layout and display issues across devices when I was building Home Value Optimizer. I’m not sure I achieved that as smoothly when I built AVM Optimizer without AI. That alone showed me how much more efficient and flexible development can be when AI is part of the process.
Q4. Were there any moments where AI fell short — things ChatGPT couldn’t quite get right, or areas where you had to step in and fix things manually?
A: Definitely. Early on, I had to prompt ChatGPT multiple times to get the results I wanted. Often, I’d save each version of the code, test it on the site, then decide what to keep or refine. It was a process of trial and improvement — but over time, the work built up and got better with each round until I was fully satisfied.
One big advantage of using ChatGPT was that it could create and edit entire sections of code at once, not just small snippets. That allowed me to see the full result in context — something that would be difficult for a traditional human programmer to do as quickly. It made the whole process much easier and more fluid.
AI is powerful, but it still needs the original idea and vision from a human. That’s where the AI-human partnership becomes so valuable. You have to guide it — tell it what you want, test what it gives back, and keep shaping the results. The best part, though, is that ChatGPT never gets tired or frustrated, no matter how many revisions or ideas you throw at it.
Q5. You’ve mentioned before that Home Value Optimizer can make AVMs up to 30% more accurate. How does AI help make that possible?
A: The concept behind Home Value Optimizer is actually quite simple — but it’s the ease of use, functionality, and clarity of results that make all the difference. As noted on the About page, the valuation results for both the AI-built HomeValueOptimizer.com and the original AVMOptimizer.com are identical.
Where ChatGPT really helps Home Value Optimizer excel is in how it structures and presents the process. It made the interface more intuitive, improved the flow of information, and made the entire experience clearer for users. That combination of simplicity and precision is what makes it effective — and that’s where AI’s influence really shines.
Q6. What advice would you give to other professionals or entrepreneurs who want to use ChatGPT or AI to build their own tools or businesses?
A: The most important thing is to jump right in and not be afraid to experiment. Try things over and over — make big changes, small changes — and see what happens. That’s how you learn what works best.
Ask ChatGPT for suggestions, bounce ideas off her — she never gets tired or frustrated. The more you engage and iterate, the better the results become. It’s a true creative partnership, and the only real way to get the most from AI is to start using it and keep refining as you go.
Q7. What’s next for Home Value Optimizer — and how do you see AI shaping the future of real estate valuation?
A: I think we’ll start seeing more consumers doing their own home valuations — much like how people can now choose to do their own taxes with software. In many cases, certain homes and market areas could actually be valued more accurately and much faster through these AI-assisted tools, with human appraisers focusing on only the most complex properties.
That said, humans will always play a key role — especially when it comes to assessing condition and upgrades. Whether it’s homeowners, inspectors, or appraisers, the human eye and experience still matter. Interestingly, I can see a future where some home inspectors become even better equipped to value properties using tools like Home Value Optimizer or AVM Optimizer than traditional realtors or appraisers.
AI is going to change a lot about how real estate works, and my goal with both Home Value Optimizer and AVM Optimizer is to help the industry adapt and improve — combining technology and human expertise to make home valuation more accurate, transparent, and accessible for everyone.
Closing Note
To see the AI-powered tool in action, visit HomeValueOptimizer.com — or explore its original human-coded counterpart at AVMOptimizer.com.
Thank you for posting this interesting and informative Q& A. Having been a licensed residential real estate appraiser for 13 years I would never have imagined what is happening now. Anything to make evaluations of properties easier and more accurate is a good thing in my opinion. I don’t think anyone ever wants to see 2008/2009 again.