His team planned a six-country leadership tour in four weeks

Ask an event professional how much lead time a global event series needs, and you'll hear eight to twelve months.

Peter Bordes' team at Circle just executed a leadership tour across six countries and three continents with four weeks of lead time.

Peter runs events at Circle, the company behind the USDC stablecoin and the Layer 1 blockchain Arc designed for the agentic economy, where the calendar runs 150 to 200 events a year, most of which are field marketing plus a set of large corporate-hosted events. August is the team's only slow month, and even that includes a major event in Latin America.

His path into the seat is a story on its own. He started in association events right out of college, and since associations run on small budgets, nobody wanted to pay for a dedicated salesperson. Peter knew the event offering deeper than any salesperson could, so he did both jobs. A chance introduction pulled him into crypto over a decade ago, and along the way he booked Circle executives, including the current CEO, as speakers at his own events, and spent years hassling their director of field marketing to sponsor them. When Circle hired him, that same director ended up reporting to him.

What makes Circle unusual isn't that leadership tells everyone to use AI. It's that leadership actually uses it, from the CEO down, and shows the work. The company also has a business reason to care: if an agent makes a phone call on your behalf and it costs five cents, someone needs to move five cents. Stablecoins are built for exactly that, so the events team sits inside a company betting on the agentic economy it plans events about.

Here's what that looks like on the events team, in practice.

Every Circle event now ships with its own chatbot. You know the master plan document with every answer an attendee could want? Nobody reads it, and everyone texts the event manager anyway. At Circle's company offsite a month ago, about 980 people, the chatbot absorbed the what-do-I-wear and what's-the-climate questions almost entirely. What still reaches a human: emergencies and the occasional room-charge dispute. Peter's verdict: "Such a lifesaver." They now do this for every event going forward.

A dashboard for deciding which events are worth it. When you run 200 events a year, the hard question isn't how to execute, it's which ones deserve the budget. The team is building a dashboard where you feed it an event's website and a few parameters, and it scores the event against Circle's objectives and tells you why. If the answer is yes, it generates the event brief and lays out the timeline with the internal stakeholders already tagged. The goal is to have it operational and shared across the company in the second half of the year.

An internal AI marketplace, where employees publish apps and skills for each other. Peter's favorite aggregates every piece of developer feedback Circle collects, categorized by region and experience level. When he designs a developer event, he invokes it and reads what his exact audience has been saying.

And his boss, the CMO, built a digital twin of himself. Anyone with a brief, a campaign, or an idea runs it through the twin first and gets feedback as if they'd met with him. Peter calls it "pretty damn accurate," and the effect is real: "It saves meetings, it saves conversations."

A few months ago, Circle's AI labs team ran a hackathon with no humans in it at all, only agents. Peter's next experiment is pointed at our industry: an agentic event. He wants to have agents design one session of his team offsite, no human input, then compare the agenda a team of agents builds against the one his human team builds. "I'll probably start with just one session to be cautious." He pegs the full version as a 2027 project.

I pushed back a little on agents, because I'm still skeptical of letting them make decisions without guardrails, and Peter isn't reckless about it either. His excitement is about something more grounded:

That's the whole thesis of this newsletter in one sentence. The four week tour wasn't magic. It's what becomes normal when leadership models the tools, knowledge gets packaged into skills anyone can invoke, and every event ships with its own AI layer.

You can find Peter on LinkedIn.

Till next time,
Noah

Do More With Less Using AI

PS: What's the first task you'd hand an agent for your next event?