Phil Stringer's Success Summit keynote explains how AI agents work like employees, not tools, and what real estate teams need to start.
At the 2026 Ylopo Success Summit in Charleston, AI strategist Phil Stringer opened his keynote with a story about an employee who spent four months and $12,800 sorting through 28,000 customer service emails. The task: pull out every question a client had ever asked, organize it by frequency, and draft suggested responses.
Phil did the same job with 26,000 emails last week. It took ten minutes and cost nothing.
One AI agent, working alone, produced a cleaner result than a full-time employee working on the project for a season.
That gap, four months versus ten minutes, is most of the talk in miniature. Something changed in how AI can be used, and a lot of people building real estate businesses haven't caught up to it yet.
That catching-up problem starts with a basic misread of what AI actually is right now.The shift from tool to team
For the last few years, most people have used AI the way Phil used to teach it himself: as a tool. Type a request into ChatGPT, it responds, refine the request, it responds again. A back-and-forth, useful but limited to whatever someone is willing to sit and type.
Phil's argument is that this era is ending.
AI agents don't wait for a prompt every time. They think, plan, and take action on their own, based on goals and information given to them ahead of time.
Instead of a tool a person operates, an agent behaves more like an employee to manage, one that can work overnight.
He drew a line between three levels of automation. Traditional automation follows fixed rules: if this happens, do that, nothing more. It doesn't learn or adapt.
AI automation adds a layer of pattern recognition on top of those rules, so the system gets a little sharper over time. AI agents go further.
They make decisions and carry out multi-step actions without a person clicking "go" at every stage.
Most real estate professionals, in Phil's read of the room, are still operating at the first level or typing into the second. The real leverage he's describing, the kind that turns four months into ten minutes, sits at the third.That leverage doesn't come from a smarter chatbot. It comes from what the agent is actually built out of.
What an agent actually needs
Phil broke an AI agent into three components, the same way someone might think about hiring a person for a role:
A brain: the large language model doing the thinking. Someone could choose ChatGPT, Claude, Gemini, or Perplexity as the brain, and swap between them later.
Memory: without it, an agent starts over every conversation and never gets better at understanding the business behind it. How much it remembers is a choice, anywhere from a few recent exchanges to hundreds of pages of history.
Tools: this is the piece Phil says most people skip, and it's the one that turns a chatbot into an employee. An agent with a brain and memory but no access to an inbox, calendar, or CRM can only talk about the business.
Connect it to Gmail, Follow Up Boss, Google Drive, or Canva, and it starts acting inside the business.
Phil was blunt about the most common mistake he sees: people set up a brain and memory, skip the tools, and then wonder why the agent can't actually do anything for them.
From there, the talk moved past the individual agent and into what a team of them starts to look like.
Building an AI team, not just an AI habit
The practical center of the keynote was an AI "operating system," Phil's term for it, a folder living on someone's own computer that holds a job description and a memory file for every AI employee they've hired. A CFO agent, a marketing director, each one scoped to a single role, the way a team wouldn't hand finance and marketing to one generalist at once.
His reasoning: a generalist prompt tends to get a generalist answer. An agent with a specific job description, written the way someone would write one for a new hire, tends to produce sharper, more consistent output.
And because the whole system lives as files on a personal computer rather than locked inside one platform, any AI tool can be pointed at that folder and pick up the context immediately. If a stronger platform shows up next month, the work already put in doesn't disappear with it.
For anyone starting from zero, Phil's advice ran simple: pick one platform, connect a couple of tools to it, then hire a first agent, usually an executive assistant, followed by a marketing director. Not the full operating system on day one.
Just one hire, and see how much it takes off the plate.
That advice landed differently live, with Phil building the actual folder structure on screen while questions came in from the room. Reading it secondhand is one thing. Watching him do it is another.
Watch the full session
Phil walked the room through live demos of all of this: importing years of ChatGPT memory into Claude in under two minutes, connecting a CRM through Zapier's MCP integration, building a custom AI skill from scratch on stage. His complete keynote from the 2026 Ylopo Success Summit is below.
Watch:
Where Ylopo fits in
The first of those 30 days, in Phil's framework, starts in the same place every time: connecting an agent to the tools a business already runs on, the CRM, the inbox, the leads coming in.
That's the same problem Ylopo AI, Ylopo's text and voice bot, was built to solve for real estate specifically.
It's an AI agent already wired to work a lead, follow up at the right time, and hand off a warm conversation to an agent, without anyone having to build those connections from scratch.
Anyone who left Phil's talk wanting an AI agent that already understands real estate lead conversion can see how Ylopo AI works.
Aaron Franklin
Head of Growth
Aaron "Kiwi" Franklin is the Head of Growth at Ylopo and a serial technologist and entrepreneur who has over 25 years of experience creating digital solutions for major brands and pioneering companies where technology and real estate meet. His depth of expertise stems from leading development of the first website for Apple to founding a global community of over 1,000 elite athletes.