Previous In a conversation with HousingWire’s Allison LaForgia, Homes.com’s Livia Sponseller and Andy Ventura discussed how Homes Ai is reshaping the online home search experience through conversational technology that benefits both consumers and real estate agents.
Livia Sponseller, Senior Vice President of Product Management at Homes.com, began by framing the feature as a way to rethink the home search process, which has remained largely unchanged for years. “What excited me most about this feature is that we were looking for a way to just completely change the way that people searched for homes online,” she said. “You’ve had the same basic map placards and standard filter methodology on these home search portals for decades.”
The goal, she explained, was to introduce a more natural, interactive approach. “It created this conversationalized, interactive home search, which we’re really excited about.”
Andy Ventura, Vice President of Enterprise Software Architect at Homes.com, noted that the technology allows users to interact with the platform more fluidly, removing the rigid steps common to traditional property searches. “It’s a different way of interacting with an already very powerful web application,” he said. “You can ask questions about grocery stores or mortgage rates or what the neighborhood is like without losing where you are in the application.”
The pair described the launch as a “generational shift” in how consumers search for homes, driven by the integration of AI directly into the browsing experience rather than treating it as a simple add-on.
“Across marketplaces . . . AI is often like a chatbot afterthought,” Ventura said. “We felt like it should be ingrained in the application. AI and the UI are both knowledgeable of each other. You talk to it, the UI responds, and when you move the map or change a filter, the AI knows the user did something.”
Sponseller said the conversational model allows buyers to explore listings more naturally and uncover insights that would normally require multiple searches and filters.
“Now the site has come officially to life,” she said. “You are talking to our website and it is talking back. You could say, ‘Hey, Homes, can you tell me why the first listing is more expensive than the second?’ and it will start laying out the differentiators.”
She described the home search process as a complex journey in which buyers try to narrow down countless options. “Your home is like a needle in a haystack that you’re trying to find,” Sponseller said. “This new search experience allows for those nuances that matter most to you.”
Ventura added that conversational AI also allows users to adjust searches dynamically. “You could say, ‘Hey, Homes, bump it up by 50k,’ and all of a sudden the search changes,” he said. “You can add a pool, remove the pool, add an elevator — whatever it is — and just have that back-and-forth conversation.”
Both leaders emphasized that the technology is designed to support real estate professionals rather than replace them. “Our company name is CoStar, and we’ve always viewed the agent as the star,” Ventura said. “What we’re trying to do is deliver that well-educated, well-informed consumer to that agent so they can get more deals done faster.”
Sponseller agreed, noting that the technology can streamline the process for both buyers and agents. “The agent is very much still relevant,” she said. “It helps them work better and more efficiently for their clients.”
Looking ahead, both leaders see significant opportunities as AI, property data and digital property scans become more integrated.
“What excites me is equipping consumers with all this data in one place in a very easy manner,” Sponseller said. “It’s helping them make more informed decisions for potentially the biggest financial investment in their lifetime.”Ventura added that deeper property insights could eventually provide even more predictive guidance for buyers. “Imagine understanding projected long-term maintenance costs or how a kitchen renovation might impact value,” he said. “It’s all about unlocking the proprietary data we have and feeding it to these models to get really good answers.”
