Build Your AI Team With Google Gemini Gems | Tony Ray Baker

AI Agent Advantage episode thumbnail with Tony Ray Baker and JMan Maneiro

Tony Ray Baker has been a realtor for 32 years, and right now he has an AI lawyer, a blog writer, a social media manager, a listing marketer, and a website analyst on call around the clock. His total payroll for that team: twenty dollars a month.

In this episode of AI Agent Advantage, the Tucson-based RE/MAX agent walks through exactly how he built his full AI team inside Google Gemini Gems, and how any agent can do the same before their next client call.

What Gems Actually Are

A Gem is a custom assistant inside Google Gemini that you train once and reuse forever. Unlike a standard AI chat, where every conversation starts from scratch, a Gem remembers its job, its instructions, and its voice. Tony Ray builds one Gem per role, and that single-purpose focus is the secret. His legal review Gem is loaded with Arizona real estate contracts, fair housing law, and the NAR code of ethics. His listing marketer does nothing but listing content. Each one outperforms a do-everything assistant because it never loses the plot.

His favorite move is what he calls meta-prompting: before training a Gem, he asks Gemini to write its own optimal training prompt. The tool teaches you how to teach it, and the output quality jumps from day one.

What You Will Learn in This Episode

  • What Google Gemini Gems are and how they differ from standard AI chats
  • How to train a Gem with specific prompts so it sounds like you and follows your exact workflow
  • The meta-prompting method that gets better output from the very first use
  • Why single-purpose Gems consistently beat multi-purpose assistants on focus and accuracy
  • How Tony Ray built a legal review Gem for Arizona contracts, fair housing, and the code of ethics
  • Where to grab Tony Ray’s free listing marketing GPT and put it to work today

The Takeaway

Most agents think building an AI team requires a tech budget and a consultant. Tony Ray built his on the cheapest Gemini plan there is, one role at a time, after 32 years of doing everything the manual way. If he can hand off the busywork, so can you.

Hit play above for the full conversation, including the story of the five companies he runs with this setup.

Ready to put this AI to work?

This is what we break down every week. Grab the free AI guides, or come build it live with us inside AI-Cademy.

Get the Free AI Guides
Explore AI-Cademy

Hire JMan