Meet My Agent: How Event Profs Put Personal AI Agents to Work

Image credit: Fardived/Shutterstock

Unlike conventional AI chatbots, which generally respond to individual prompts, AI agents work independently with connected tools and information to complete a series of tasks, achieve a goal, and, in general, take action with varying degrees of human oversight.

They build reports, book travel, automate workflows, register for events—whatever their human counterpart is willing, and feels comfortable, to delegate.

The September 2026 Workday survey of nearly 6,000 workers globally found that 69% of employees who use AI had allowed it to complete an entire task with little or no input at least once.

Senior leaders were considerably further along in adoption: 88% had used agents to complete tasks, compared with 54% of individual contributors. Yet just 38% of employees said they could confidently explain what an AI agent is.

Forty-one percent of employees who use AI said they deliberately conceal that use. How employees feel about being managed by an AI agent and how human agency factors into the equation are still developing topics. 

In the meantime, professionals in every industry are experimenting with AI agents for personal and professional use, pre-made or building their own, to explore whether and how they boost productivity and integrate into their lives.

Event professionals are on the forefront of this shift, and, of course, we had to ask: What do they delegate? Where does the human judgment come in? Learnings and wins? 

Virtual Colleagues

Alex Velinov, vice president of AI and technology at mdg, has approximately 20 agents supporting different parts of his work. Most handle specific, repetitive tasks, freeing him to focus on work that requires human judgment and creativity.

When he first started building agents, Velinov gave them human names and personalities, but it became hard to remember. Now, most are named after the jobs they perform. One exception is Nina Morrison, his prompt engineering assistant, which helps him write, review and refine prompts. He also has a proofreader and fact-checker that he uses every day.

“I think of my agents as virtual colleagues who help me work faster and produce better results,” he says. “A great example is my monthly reporting. What used to take me about two hours now takes roughly 20 minutes. My agent takes the data, organizes it into presentation content, helps build the deck and identifies key insights. By the time I get it back, it’s about 80% there.”

But the final 20% is crucial, he says. “That’s where I bring in my own perspective, judgment and context. I’ll go back and forth with the agent to refine things, but ultimately, I’m responsible for the final product.”

Velinov uses Claude to build his agents and looks beyond traditional platforms for specific capabilities. 

His advice about whether you need one? Yes, if there’s a tedious task you’re performing every day. “That’s where AI delivers the most value,” he says. “It’s not about replacing the human element. It’s about spending less time on repetitive work and more time on the thinking that makes the work better.”

Follow-Up Guru

Noah Cheyer, founder of Silicon Valley Speakers, also has multiple agents but the one he uses the most focuses on follow-up.

Built in Claude Code, the agent reviews his calendar and meeting notes, recorded with Granola, identifies what he promised each person and drafts follow-up emails in Superhuman for anyone still awaiting a response. It also updates the next steps for each deal in Cheyer’s CRM.

“It never sends anything on its own,” Cheyer says. “Every draft waits for me to read it and hit send, because a follow-up with the wrong date or price is still a risk.”

The biggest lesson was that an agent is only as good as what it can see, Cheyer says. “Once it could read my meeting notes and past emails, the drafts stopped sounding generic and started sounding like me. If you’re coming off a trade show with dozens of conversations to follow up on, this is the one I’d build first.”

Chief of Staff

“My Chief of Staff acts as my true assistant,” says Joe Federbush, president and chief strategist at Evolio Marketing.

It tracks priorities, deadlines, projects, and commitments across Outlook, Teams and SharePoint, proactively surfacing what needs attention. It also coordinates follow-ups, drafts proposals, and prepares summaries and briefings. And it even checks in throughout the day via Teams chat to keep its owner on track.

“Its goal is to help me stay focused on decisions and execution rather than managing information and administrative overhead,” Federbush says. Built with Microsoft Copilot Studio, Power Automate and Power Apps, the agent was developed within Microsoft’s ecosystem, because data security was a priority when working with Evolio’s sensitive information.

Federbush’s thinking about agent architecture is evolving. “A year ago, I thought the strategy was to build a bunch of agents to own specific tasks, and now, with AI advancing so quickly, the strategy is shifting to building an ‘executive’ agent with specific playbooks and sub-agents all working in sync.” 

His favorite learning? After finding the agent’s default tendency to compliment and apologize annoying, the pendulum swung to cold and robotic, which Federbush liked even less. “So, I trained it on when to be nice, when to be regular, and even when to be sarcastic. The sarcastic part is the best because it’s appropriately surprising.”

Royal Treatment

Dahlia El Gazzar, owner of Dahlia+Agency, named her personal AI agent Zaatar, after a popular Middle Eastern herb blend. “Of course it needed a little spice,” she says.

Zaatar is a dot from OpenAI, an always-on personal AI agent that works across connected apps. Its job for El Gazzar is to “keep the thread: what matters now, what we promised, and what needs to get finished.” 

This translates to tracking unanswered emails, upcoming events, project deadlines and loose ends across email and Slack. It can also draft replies in El Gazzar’s voice for her to review and approve, while helping her stay focused when competing ideas and projects threaten to pull her in too many directions.

“The biggest surprise is the persistence,” she says. “It keeps bringing me back to my priorities and what needs to happen next.”

For El Gazzar, the goal is to give her creativity room to run without letting her priorities get trampled. As Zaatar puts it, she wants help that can keep up with her energy and still say, “Queen, let’s finish this one first.”


Have a measurement story idea? Reach out to EM’s editor-at-large Anna Huddleston.