The interactive chatbot draws on Keller Williams’ proprietary models, systems, books and training resources to answer agents’ questions and assist with content creation.

At Inman Connect Las Vegas, July 30-Aug. 1, 2024, the noise and misinformation will be banished, all your big questions will be answered, and new business opportunities will be revealed. Join us.

Keller Williams is furthering its use of artificial intelligence with the launch of a new generative AI-powered real estate assistant called “KWIQ,” the brokerage announced on Monday.

The assistant is trained in Keller Williams’ proprietary models, systems, affiliated books and training resources from Keller Williams University.

Chris Cox, Keller Williams

“We are thrilled to bring this AI-powered tool to agents and look forward to its transformative impact on their work,” Chris Cox, chief technology and digital officer at Keller Williams, said in a statement. “Keeping them ahead of the curve, KWIQ empowers agents to work smarter, faster and more effectively.”

KWIQ was launched during the second quarter of 2024 as an interactive chatbot within KW’s proprietary technologies, including Command, the firm’s cloud-based, integrated CRM-plus platform, and Connect, the firm’s learning management system.

The chatbot can provide agents with information on real estate topics, assist in content creation (with social media, newsletters, listing description, etc.), and provide agents with guidance on using AI tools and automation to streamline their business processes, in practice, making it Keller Williams’ answer to ChatGPT. Unlike ChatGPT, which can be fed with misinformation from the internet, KWIQ is a closed-loop system that draws on the vetted training materials and resources that Keller Williams puts into it, Executive Director of Labs David Voorhees told Inman, which eliminates the threat of agents receiving bad information.

David Voorhees, Keller Williams

“Agents are only getting back stuff that we’ve put into it, so it can’t give them misinformation or bad information or something that’s not going to make them successful,” Voorhees said. “It’s all coming from our proven systems and models.”

The chatbot has evolved from years of Keller Williams testing out various AI resources, most recently a digital ads platform where KW agents can create digital advertising on social media, Google and elsewhere. With that tool, KW integrated ChatGPT to help agents create prompts for the tool, and that process helped inform KWIQ.

“When we polled agents that weren’t using the tool or that were hesitant, they all kind of told us one thing: I don’t know what to say,” Voorhees said. “So we added ChatGPT into our digital ads platform to help agents create content. And that’s kind of flowed through to KWIQ as one of the top things we’re seeing our agents ask about, helping them create marketing content, social media content and listing descriptions and email content. It still holds true: Agents know how to sell houses, but they sometimes don’t know how to advertise.”

Keller Williams will continue to update KWIQ so that it will ultimately be able to help agents with real-time market research and analysis. In about the next six to nine months, the company said it expects the chatbot to be able to provide data on local market conditions and trends that will help agents in advising their clients on homebuying and selling strategies.

While KWIQ is still in its early renditions, Voorhees noted that agents are able to provide feedback on the chatbot in the form of a thumbs up or thumbs down. So far, fewer than 3 percent of agents have given it a thumbs down. Right now, the tool is available to agents through Command and Connect on desktop, but soon it will also be available in a mobile version.

“This is just the beginning for KWIQ,” Cox added. “Given our advantage of having the most data in our industry due to our size, we will continue to connect market- and production-based datasets into KWIQ to aid our agents in providing the best client experience.”

Email Lillian Dickerson

This post was originally published on this site