AI Agent Configures Full Events and Gets Instant Quotes — No Human Involved

Event tech company confirms that AI agents can autonomously use its web-based configurators to build custom events and generate detailed quotes — and that the AI's planning logic goes beyond what most human users would deliver.

What happens when you let an AI agent loose on a web-based event configurator — not as a chatbot answering questions, but as an autonomous user configuring a real event?

Eventmachine, provider of automated event quoting software for hotels and venues, set out to answer that question. Over multiple test sessions, the company had Claude, an AI model by Anthropic acting as an autonomous agent, use Eventmachine's publicly available event configurators to plan corporate events and private celebrations from scratch — selecting event spaces, meeting rooms, seating layouts, catering, accommodation, activities, and technical equipment. At the end of each configuration, the AI submitted the quote request, and a fully calculated, professionally formatted quote was instantly delivered by email — just as it would be for any human user.

The AI completed the entire process autonomously. No human guided it through the interface. No custom integration was required. The AI used the same web-based configurator that event planners, corporate bookers, and hotel sales teams use every day.

"Our configurator UI is built to be logical and explicit — every option, every price, every dependency is right there on screen," says Peter Warren, CTO of Eventmachine. "We knew that if an AI could handle the interface, the results would be strong. What caught us off guard was how the AI reasoned about its choices."

What the AI Actually Did

Across multiple configuration sessions — corporate meetings, team offsites, weddings — the AI didn't just fill in fields. It made planning decisions that experienced event managers would recognize as genuinely thoughtful.

In one session, the AI selected a larger, more expensive meeting room over a smaller one that technically met the headcount requirement. The reason: the configurator showed that the larger room's hire fee would be waived if catering spend exceeded a certain threshold. The AI calculated that the planned food and beverage items would comfortably clear that threshold — making the bigger room effectively free. That's the kind of cross-reference that takes an experienced banquet manager a few minutes with a calculator. The AI spotted it while configuring.

In another, the AI split a 26-person group into two parallel afternoon activities — a golf lesson and a patisserie course — with exactly 13 guests in each stream. The choice wasn't random. Golf is active and outdoor, patisserie is indoor and sensory. The two groups reconvene at dinner with different experiences to talk about. That's not configuration — that's event design.

For a 35-person corporate kick-off, the event brief specified an activity for 20 participants. The AI looked at the activity options in the configurator and noticed that some included an input field for the number of attendees, while others did not. It reasoned that selecting an option without an attendee field would charge the full group of 35 — overpaying for an activity only 20 people would join. So it specifically chose an option with an adjustable attendee count and set it to 20. It read the interface, understood the billing logic, and made the economical choice.

"The AI treated our configurator the way a sharp, experienced planner would," says Warren. "It read room descriptions for atmosphere cues, compared pricing mechanics across different options, and figured out which selections would save money without compromising the event. All of that from information that's right there in the UI — it just connected the dots faster than most users do."

Why This Matters for Hotels and Venues

AI agents are no longer a future scenario. According to PwC, 79% of organizations have already adopted AI agents in some form. Gartner projects that by 2026, 40% of enterprise applications will embed task-specific AI agents. Google's AI Mode — which enables users to delegate complex tasks to AI — is already live in over 200 countries.

The practical implication for hospitality is straightforward: when AI agents start sourcing events on behalf of corporate travel managers, executive assistants, or wedding couples, they will need structured tools they can interact with to compare venues and obtain quotes. A venue that offers a web-based configurator with instant quoting gives the AI something to work with. A venue that requires an RFP submission and a 48-hour wait does not.

"This isn't about replacing the event sales team," says Florian Zelfel, CEO of Eventmachine. "It's about being open for business when the inquiry arrives — whether that inquiry comes from a person or an AI doing the research for them. If a prospect's AI hits a contact form instead of a configurator, the venue is already behind."

The Bigger Picture

HFTP, a global hospitality technology association, has been tracking this shift closely. In a widely cited 2026 analysis, the organization described a near-future where guest-side AI agents interact directly with hotel-side systems — requiring what HFTP called "machine-readable inventory, availability, offers, policies, fees, experiences, and upsells." Hotels lacking this kind of structured, machine-readable detail, the analysis warned, risk falling out of the candidate set entirely when AI systems source the best match for a traveler's request.

For Eventmachine, the finding validated an approach the company has long advocated: building event sales systems that are structured, standardized, and explicit. Eventmachine's configurators were designed for human users — hotel sales teams and event planners looking for a faster way to build and price custom events. But because the system was built on clear logic, transparent pricing, and well-organized options rather than free-text forms and manual processes, AI agents can use it too.

"We've spent years telling our clients: get organized, standardize your offerings, make your pricing logical and transparent," says Zelfel. "That discipline is what made this possible. The AI didn't succeed because we got lucky — it succeeded because the system was built right."