Gramercy Tech won Best AI for Attendee Engagement, Networking & Matchmaking for Seek, an AI-powered activation that turned an entire conference floor into a playing field, with Google Gemini as the real-time judge.
Every experienced event professional knows the quiet problem with large internal conferences: programming is passive, networking is largely accidental, and energy drops off sharply after day one no matter how strong the content is. People show up, sit in sessions, make polite conversation at dinner, and leave having met roughly the people they already knew. For the UPS Leadership Conference, Gramercy Tech was asked to fix that, without a dedicated operations team running it in real time, and in a way that felt like UPS rather than a branded novelty bolted on top.
Their answer, Seek, won the AI Edition’s attendee-engagement category by doing something the judges rated as genuinely different: it made the AI the active engine of the experience, not a supporting feature within it.
A playing field, not a platform
Seek is an AI-powered scavenger hunt, delivered through Gramercy Tech’s GXP ecosystem inside the existing event app, so there is no download friction and no separate platform to learn. Attendees earn points by photographing items and completing tasks from a curated daily list. Google Gemini validates and scores every submission in real time, with no human in the loop, and when a submission falls short it tells the attendee exactly why, in language specific to what they sent. A live leaderboard runs from opening session to closing keynote.
The rigour underneath
This is where most competitors cannot follow. Standard AI validation broke in testing in instructive ways: people photographed screens instead of physical objects, or images that met the letter of a task but not its spirit. So Gramercy Tech built a multi-layered validation architecture on top of Gemini: proprietary prompt frameworks, task-specific parameters, and exhaustive QA across valid and invalid submissions before a single attendee touched it. Gemini’s role is visible and active: it identifies brand colours, reads attendee badges via high-velocity OCR to verify cross-regional connections, and applies spatial reasoning to estimate package dimensions. The standard the team held itself to was not “mostly works” but zero manual adjudication across a live, multi-day event. That is what was delivered.
Culture, embedded in the tasks
The creative dimension is the task design. Prompts were mapped to the specific physical and cultural landscape of the event: finding a particular shade of UPS brown somewhere on the floor, scanning a napkin at the gala dinner, connecting with a colleague from another region and capturing badge proof of the encounter. Every prompt was a deliberate decision informed by UPS’s culture, the venue layout and the programme. That list was only possible because Gramercy Tech understood both what the platform could push and what UPS’s people would actually respond to, a collaboration that produced something neither side could have built alone.
The results
35% of attendees actively participated, the highest the activation had ever achieved and a 25% year-on-year increase, in a version that asked considerably more of attendees than any before it. For a voluntary, gamified activation at a senior corporate audience, that is a meaningful behavioural shift. UPS leadership took part personally, which at that level signals organisational trust and visibly changes the energy in the room. Afterwards, attendees consistently pointed to three things: discovering corners of the venue they had no other reason to visit, conversations that would not have happened without a shared task, and a felt connection to UPS’s values that never registered as branded content, the hardest outcome to engineer and the most telling measure that it worked.
The judges saw it clearly. Pietro Landini called it “a compelling and well-evidenced entry,” praising a “genuinely innovative use of AI in a live event context, moving beyond passive validation to position Gemini as an active, real-time participant.” Shervin Mashayekh described “a well-executed and engaging use of AI” that is “practical, accessible, and clearly connected to the live event experience.”
“A genuinely innovative use of AI in a live event context, moving beyond passive validation to position Gemini as an active, real-time participant in the experience.”
— Pietro Landini, Judge, Event Technology Awards AI Edition 2026
Why it matters for the industry
AI in live events is producing enormous creative possibility and enormous overpromising at the same time. Activations that shine in a controlled demo fail when real people show up in real numbers, because they were never stress-tested at scale, never given custom safeguards, never designed around the specific audience in the specific room. Seek was built to sit on the right side of that gap. It makes the case that AI, designed with intention and built with genuine technical rigour, can be the central engine of a live experience rather than a supporting feature, exactly the standard this category exists to reward.
About Gramercy Tech – Founded in New York in 2003, Gramercy Tech is an experiential event technology company that builds its platform and its team together, across Event Management and Event Experiences (GXP). It runs on a 100% proprietary, SOC 2 Type II certified platform rebuilt in 2023 with an AI-first architecture, and has delivered more than 20,000 events for clients including Google, Walmart, PwC, Deloitte and the X Games. gramercytech.com
